Volume 274 Number 1151January 2015REVIEWS AND COMMENTA - PDF document

1 Volume 274: Number 1—January 2015RE
Volume 274: Number 1—January 2015REVIEWS AND COMMENTA Published online10.1148/radiol.14142265Content code: Radiology 2015;From the Editorial Ofce, 800 Boylston St, 15th Floor, Boston, MA 02199. Received September 23, 2014; nal version accepted September 23. Deborah Levine, MDHerbert Y. Kressel, MDRadiology’s Legacy: through the Atticucked away in hallways and corners of the Radiology Editorial Figure 1 Figure 1:media, with striations in the middle of the defect due the villous protrusions. Gross pathologic appear Volume 274: Number 1—January 2015 Figure 3 Figure 3: Radiograph shows stones grouped in the urinary bladder in 60 mm (g 1 from reference 3). A total of 25 faceted stones were removed (g 2 from reference 3). Figure 2 Figure 2:These early articles are based on conventional radiography and provide the foundation for both the way we think about extracting diagnostically important information from images and the spectacular technologic achievements in the decades that followed. Thus, we decided to include in this Radiology a pictorial essay of select gures and our thoughts on the articles themselves to give a taste of the beginnings of so many of our current imaging and research activities.The rst salient observation when reading the early works in Radiology is that the early articles were predominately descriptive (Figs 1–3) (1–3) and relied heavily on personal experience. There was commonly no description of the study methods, no control groups, and very little, if any, quantitative analysis. Early radiologists shared their observations obtained through careful scrutiny of images (Fig 4(4) and comparison of their observations with those reported previously by others. In fact, many of these early Figure 4 Figure 4:Diagram drawn over a normal heart in a child aged 9 years obtained at a distance of 6 feet. = descending aorta, = great vessels, = left auricle, = left ventricle, left diameter, = middle right diameter, pulmonary artery, =, right auricle; ventricle. Middle right diameter (3.5 cm) plus middle left diameter (7 cm) equals 10.5 cm, which is half Figure 5 Figure 5:Radiograph in an 18-month-old child with lead poisoning. Note the bands of increased opacity at the distal ends of the radii, ulna, and Volume 274: Number 1—January 2015 Figure 7 Figure 7: Radiograph shows an open safety pin in the upper third of the esophagus with the point down (g 1 from reference 8). Photograph of the patient 2 hours after removal of the safety pin (g 2 contributions read like review articles until the end of the report, when the new information and the authors’ personal experience and views are shared with the reader. Correlation with gross appearance, function, and pathologic ndings were common. Caffey’s classic description of lead lines in 1931 (Fig 5is one such example (5). Commentary on articles was also frequently incorporated into the text. For example, the importance of a personal interaction between experts in pathology and radiology was encouraged. Dr Robertson wrote, “The pathologist is usually more than willing to cooperate. Will not the roentgenologist fall short of his duty and privilege if he does not seize these opportunities? We, as physicians, owe it to the living that our deaths shall not Original articles were the spring point for many of the named abnormalities that we use in general practice today. For example, the 1940 article by Dr Hill and Dr Sachs entitled The Grooved Defect of the Humeral Head: A Frequently Unrecognized Complication of Dislocations of the was the rst published description of what we now term the The focus of these articles was typout sophisticated statistical analysis. It is evident from perusing these articles that the questions being answered were pertinent to patient care of the times (Figs 8–10) (9–11). A common disease of the day, tuberculosis, was featured frequently, with commentary on both Figure 8 Figure 8:Radiograph shows a cavity in a patient with tuberculosis. Although there has been considerable destruction in this patient, the tendency to form brous tissue, as evidenced by the rm wall to the cavity, the bands of opaque shading in each upper lobe, and calcication in each hilum lead one to conclude that repair is occurring far more rapidly than destruction. Thus, a favorable prognosis was Figure 6 Figure 6:After correction of the third dislocation, a large groove is visible when the humerus is in marked internal rotation (g 3 from reference 7). Volume 274: Number 1—January 2015the time. It was only over time that practitioners and journals adopted a more standardized approach toward presentation of images.Structured quality assurance and peer review were not yet a part of imaging departments; however, in these early articles, authors commonly shared their personal experiences with pitfalls in diagnoses. Remarkably, the potential harms of overdiagnosis or an overly pessimistic prognosis based on imaging ndings were emphasized (Fig 8) (9).Radiology has always portant and classic observations on the behavior of x-rays were publis

2 hed in the journal (Fig 11) (12). For ex
hed in the journal (Fig 11) (12). For example, the article by Compton (12), described his work that eventually led to a Nobel prize. Safety issues were of concern, with attempts being made to standardize the radiation dose. There were many articles on radiation physics, use of Bucky diaphragms, and spot lms. Figure 12(13) shows some of these aspects of early publications.While there was interest in developing a better understanding of the Figure 9 Figure 9:Radiograph shows chronic miliary tuberculosis. Note the left-right inversion of the image, as standardized projection orientation was not yet estab Figure 10 Figure 10:Radiograph shows bone changes due to absorption without amputation. The most notable changes occur in the phalanges of both extremities and in the nasal septum, whereas the bones of the trunk and the long bones of the limbs are unaffected. Like other symptoms of leprosy, there is often a curious limitation to the most peripheral parts. The time factor involved in the loss of bone is widely variable, but most often the process is very slow, with 10–30 years passing while the digits are gradually resorbed diagnosis and treatment of this disease. What is a bit surprising when viewing dardization of the black on white (Fig 8) (9) and even the left-right (Fig 9) (10) view of the images, with different viewing techniques being used even within individual articles. No doubt this large degree of variation in techniques characterized the state of practice at Figure 11 Figure 11:When an x-ray quantum is scattered by an electron at an angle , the electron recoils at an angle , thereby removing some energy from the quantum and reducing its Volume 274: Number 1—January 2015 Figure 12 Figure 13 Figure 13:The relative size of the fetal head and the mother’s pelvis is better seen in the anteroposterior view. The article states: “The amount of radiation required to make two or three exposures of the abdomen is well below the minimum dosage which would have any injurious biological effect on mother or fetus. As a matter of fact, the x-ray department of the Magee Hospital has made all sorts of roentgen examinations on pregnant women daily for the last ten years, and we have yet to see the rst case in which there is any suspicion that such examination has done harm” use of x-rays in imaging and therapy, there was a noticeable lack of understanding of the dangers of radiation used for diagnostic imaging. For example, in descriptions of obstetric imaging, there was recognition of potential harm from radiation exposure; however, this tended to be underemphasized, since it was believed that harm would be minimal if only a few images were obtained. Specically, in the article from which Figure 13 was taken, Grier states, “The amount of radiation required to make two or three exposures of the abdomen is well below the minimum dosage which would have any injurious biological effect on mother or fetus. As a matter of fact, the X-ray department of the Magee Hospital has made all sorts of roentgen examinations on pregnant women daily for the last ten years, and we have yet to see the rst case in which there is any suspicion that such examination has done harm” (14). Such a quote clearly demonstrates the importance of population studies and statistical analysis rather than the anecdotal reporting of experience.Descriptions of a technique, such as how to obtain dental x-rays, published in 1925, reveal a remarkable similarity to techniques used today ) (15). The imaging apparatus has changed over the years, a movement from analog to digital images has occurred, and overall radiation dose has decreased; however, given the many improvements in hardware and software and a better understanding of exposure and voltage, the key information provided by the images remains similar to the information provided alA noteworthy early development was the use of contrast agents to improve the scope of anatomic detail Figure 12: Diagram of a working Circuit crest meter developed by Dr Rieber to improve methods of measuring electrical input to x-ray tubes (g 1 from reference 13). Photograph of Figure 14 Figure 14:radiograph (g 4 from Volume 274: Number 1—January 2015visualized or, equally importantly, to assess function. One can only imagine the excitement felt by early radiologists Figure 16 Figure 16:Radiograph of a normal knee joint. The suprapatellar pouch normally assumes the shape of a gall bladder. The posterior compartment is elliptical in shape. The infrapatellar space is usually triangular in shape, with the apex at the patella and the base slightly irregular due to the position of the cartilage. , and were lled with carbon dioxide Figure 17 Figure 17: Radiograph shows the right hydrosalpinx contains a “pearl cluster of lipiodol.” tory diagram. The article states, “With our increased experience in the use of combined iodized oil and pneumoperitoneum in gynecology, we feel that we may recommend this method as a safe and precise means of pelvic diagnosis” (g 1 from reference 18). Figure 15 Figure 15: Anteroposterior image in a patient wit

3 h epilepsy and marked mental decien
h epilepsy and marked mental deciency shows bilateral atrophy in the island of Reil areas, poor lling of the cortical air spaces, and only partial drainage of the right ventricle (g 7 from reference 16). Photograph shows the position of the patient to obtain an anteroposterior ventriculogram with the Bucky diaphragm tilted forward, the head of the patient xed, and the plate changer and the body in position for this exposure. Note that the height of the chair permits the position of the patient to be changed without disturbing the height of the Bucky diaphragm. The Bucky magnetic shift is operated from the transformer control (g 2 from reference 16).and that physiologic function could be evaluated. Many different types of contrast, ranging from negative contrast from air or gas to positive contrast with iodine, and methods of contrast application were described (Figs ) (16–20). As stated in an article published in 1929, “Startling revelations have been made in almost every branch of medicine and surgery in the past decade, but few have attracted the attention of the medical profession as has the use of a contrast medium in roentgen diagnosis….the iodized oils have almost unlimited uses” (20). Furthermore, an article published in 1930 stated, “Intravenous urography will undoubtedly solve many urological problems without the aid of cystoscopy and ureteral catheterization, and will probably replace catheter pyelography as the method of choice” (19). These articles described the thought process behind the use of contrast agents, described the manner in which contrast when they realized that the urinary tract and gastrointestinal tract could in fact be visualized radiographically Volume 274: Number 1—January 2015 Figure 18 Figure 18:(a, b) Note the variation in appearance of the urogram at the second () and third ) examinations. In , the ureters are in systole Figure 19 Figure 19:Injected maxillary antrum reveals Figure 20 Figure 20:Radiograph of the right leg shows fracture. Exposure was made through the casket. The article states, “During the St. Louis cyclone of September, 1927, a workman for one of the large factories was killed. Eighteen months later, the exact cause of death was a disputed question. An autopsy committee was appointed to determine whether death had been caused by direct violence or by burns from bursting steam pipes. The condition of the body at the time of death prevented thorough embalming. Upon exhumation, it was found that the poor state of preservation prevented the body’s removal from the casket.… The only pathology from a radiographic standpoint was a compound fracture through the middle of the right tibia and bula. This examination was conducted in the eldhouse of the cemetery, with the thermometer unfortunately standing close to zero” (g 1 agents were administered, and showed both normal and abnormal ndings. We radiologists take many of these concepts for granted today in the era of cross-sectional imaging.We often think of early medical radiography being performed with large cumbersome devices that were available Figure 21 Figure 21:Radiograph shows the retention of two drams of water and barium mixture, with the lower level of the meal being opposite the fourth dorsal vertebra. A diagnosis of complete occlusion of the esophagus was made. The following day, the infant died. An autopsy showed that the barium ched the stomach; however, the pouch as shown on the radiograph did not have an outlet. On passing a soft probe from the stomach upwards, it was found that there was an opening into the trachea at the level of the fth dorsal vertebra and that the barium had reached the stomach by passing down through the vocal cords and the upper part of the trachea through a congenitally formed passage into the lower part of the esophagus and then into the stomach (g 1 facilities. However, articles in the 1920s showed use of portable x-ray machines in the diagnositic process (Fig 20(21) and as a forensic tool in the determination of cause of death (Fig 21(22). Remote diagnosis, an advance popularized in the late 1990s, was described in 1925 in a short two-page description of transmission of soft-copy radiographs (Fig 22) (23). In the aforementioned article, Guy Bartlett of the General Electric Company described a scenario where “an x-ray lm is made of the injury, sent by wire immediately to specialist in a distant city, and within an hour an answer is received by telegraph.” The article Volume 274: Number 1—January 2015 Figure 22 Figure 22:: Photograph of original radiograph obtained in New York, NY. Right: Photograph of the same image printed from an unretouched negative, as received by wire in Chicago, Ill. This ushered in the future of medical consultations. Details were not lost in the procedure, and an accurate analysis of the image was possible. The picture received in Chicago was immediately sent to New York by airplane mail and was published on the following morning by one of the New York newspapers (g 1 Figure 23 Figure 23:Examination of the right eye showed a

4 perforating wound of the sclera on the n
perforating wound of the sclera on the nasal side about 1.5 mm from the limbus. Radiograph shows a large foreign body 9 × 2 × 4 mm in the vitreous chamber, with one extremity close to the sclera of the temporal side and localization of this fragment so that scleral incision could be made over one end of it, thereby enabling its Figure 24 Figure 24: Radiograph shows a normal stomach and duodenum in normal position (g 18 from reference 24). Radiograph shows the right lateral aspect of the stomach. The pylorus and duodenum are displaced forward. The interpretation was that of a retroperitoneal tumor, apparently arising from the Volume 274: Number 1—January 2015goes on to quote Dr Steiner, a Radiologist at Roosevelt Hospital in New York, NY, as stating, “Medical consultation between doctors on X-ray plates transmitted by radio or telegraph is a realization that should nd a most useful place in these progressive times” (23). When we think back to the technology available at that time, it is clear that the ability of the telegraph to transmit photographs for newspapers was the basis for this discovery.Cross-sectional imaging had not yet been introduced, but deductive reasoning allowed for determination of the location of foreign bodies and tumors that could not be elucidated by physical examination (Figs 23, 24(24,25). When we look at where imaging is today, with ultrasonography, computed tomography, magnetic resonance imaging, positron emission tomography, and various combinations of these modalities, it is clear that our specialty has benetted from those early descriptions and has advanced rapidly with new technology as it has become available. Diagnosis in the past was limited to the tools available at the time. The number of available imaging tools has expanded dramatically, and our analysis tools have increased in their sophistication. We hope our readers will enjoy this sampling of the rich heritage of imaging research and that they will be enticed to discover for themselves the legacy of RadiologyDisclosures of Conicts of Interest: D.L. Activities related to the present article: none to disclose. Activities not related to the present article: receives royalties from Elsevier; is an editor for UpToDate and Amyrsis. Other relationships: none to disclose. H.Y. K. Activities related to the present article: none to disclose. Activities not related to the present article: receives royalties from Medrad/Bayer-Schering. Other relationships: none to disclose.ReferencesIngber IS. Papillomatous growths of the stomach: a report of two cases. Radiology Meyerding H. X-Ray ndings in bone tumors: exostosis, chondromas, bone cysts, osteitis brosa cystica, giant cell tumors. Radiology 1924;3(3):216–221.3.O’Hara FS. Two cases of faceted calculi of the urinary bladder. Radiology 1925;5(1):72–Perkins C. Roentgen cardiac studies in children. Radiology 1924;2(6):367–373.Caffey J. Clinical and experimental lead poisoning: some roentgenologic and anatomic changes in growing bones. Radiology 1931;Robertson HE. The signicance of postmortem examinations in the training of roentgenologists. Radiology 1928;11(6):493–497.7.Hill H, Sachs M. The grooved defect of the humeral head: a frequently unrecognized complication of dislocations of the shoulder joint. Radiology 1940;35(6):690–700.Vinson P. Unusual position of an open safety pin in the esophagus. Radiology 1925;4(2):MacRae JD, Frazer T. Roentgenographic pessimism. Radiology 1923;1(4):235–237. 10.Bierman M. Chronic miliary tuberculosis and healed miliary tuberculosis. Radiology 1926;11.Hopkins R. Bone changes in leprosy. Radiology 1928;11(6):470–473.Compton AH. Scattering of X-rays. RadiolRieber F. Standardization of Roentgen output. Radiology 1923;1(3):153–160.Grier GW. The value of the lateral view in the diagnosis of pregnancy. Radiology 1930;14(6):571–575.Blue JA. Department of Radiodontia: the advantage of rigid lms of various sizes in radiographing the mouth. Radiology 1925;5(4):Pancoast KH, Fay T. Encephalography as the roentgenologist should understand it: an attempt to standardize the procedure. Radiology 1930;15(2):173–210.17.Bernstein MA, Arens RA. Diagnostic radiological study. Radiology 1926;7(6):Stein IF, Arens RA. Visualization of the pelvic viscera iodized oil and pneumoperitoneum combined in gynecology. Radiology Hirsch JS. Urography by uroselectan. RadiFriedman LJ. Iodized oil in roentgenology. Radiology 1929;12(2):114–126.21.Titterington PF. Portable radiographic autopsy examination. Radiology 1930;15(1):Brown R. Congenital deformity of trachea and esophagus: report of case. Radiology Bartlett G. X-ray consultation of the future. Radiology 1925;5(5):439–440.Mills HP, Watkins WV. Localization of foreign bodies in or about the eye. Radiology Brown B. Differential diagnosis of abdominal tumors by the roentgenological method. Radiology 1928;10(1):48–56. SPECIAL REVIEW: Radiology’s LegacyLevine and Kressel SPECIAL REVIEW: Radiology’s LegacyLevine and Kressel 28 radiology.rsna.org  Radiology: Volume 274: Number 1—January 2015 goes on to

5 quote Dr Steiner, a Radi - ologist at R
quote Dr Steiner, a Radi - ologist at Roosevelt Hospital in New York, NY, as stating, “Medical consul - tation between doctors on X-ray plates transmitted by radio or telegraph is a realization that should nd a most use - ful place in these progressive times” (23). When we think back to the tech - nology available at that time, it is clear that the ability of the telegraph to transmit photographs for newspapers was the basis for this discovery. Cross-sectional imaging had not yet been introduced, but deductive reasoning allowed for determination of the location of foreign bodies and tumors that could not be elucidated by physical examination (Figs 23, 24 ) (24,25). When we look at where im - aging is today, with ultrasonography, computed tomography, magnetic res - onance imaging, positron emission to - mography, and various combinations of these modalities, it is clear that our specialty has benetted from those early descriptions and has advanced rapidly with new technology as it has become available. Diagnosis in the past was limited to the tools available at the time. The number of available im - aging tools has expanded dramatically, and our analysis tools have increased in their sophistication. We hope our readers will enjoy this sampling of the rich heritage of imaging research and that they will be enticed to discover for themselves the legacy of Radiology . Disclosures of Conicts of Interest: D.L. Activ - ities related to the present article: none to dis - close. Activities not related to the present arti - cle: receives royalties from Elsevier; is an editor for UpToDate and Amyrsis. Other relationships: none to disclose. H.Y. K. Activities related to the present article: none to disclose. Activities not related to the present article: receives royalties from Medrad/Bayer-Schering. Other relation - ships: none to disclose. References 1. Ingber IS. Papillomatous growths of the stomach: a report of two cases. Radiology 1923;1(1):50–52. 2. Meyerding H. X-Ray ndings in bone tu - mors: exostosis, chondromas, bone cysts, osteitis brosa cystica, giant cell tumors. Radiology 1924;3(3):216–221. 3. O’Hara FS. Two cases of faceted calculi of the urinary bladder. Radiology 1925;5(1):72– 73. 4. Perkins C. Roentgen cardiac studies in chil - dren. Radiology 1924;2(6):367–373. 5. Caffey J. Clinical and experimental lead poi - soning: some roentgenologic and anatomic changes in growing bones. Radiology 1931; 17(5):957–983. 6. Robertson HE. The signicance of postmor - tem examinations in the training of roent - genologists. Radiology 1928;11(6):493–497. 7. Hill H, Sachs M. The grooved defect of the humeral head: a frequently unrecog - nized complication of dislocations of the shoulder joint. Radiology 1940;35(6):690– 700. 8. Vinson P. Unusual position of an open safety pin in the esophagus. Radiology 1925;4(2): 129–130. 9. MacRae JD, Frazer T. Roentgenographic pessimism. Radiology 1923;1(4):235–237. 10. Bierman M. Chronic miliary tuberculosis and healed miliary tuberculosis. Radiology 1926; 6(2):165–166. 11. Hopkins R. Bone changes in leprosy. Radiol - ogy 1928;11(6):470–473. 12. Compton AH. Scattering of X-rays. Radiol - ogy 1924;3(6):479–485. 13. Rieber F. Standardization of Roentgen out - put. Radiology 1923;1(3):153–160. 14. Grier GW. The value of the lateral view in the diagnosis of pregnancy. Radiology 1930; 14(6):571–575. 15. Blue JA. Department of Radiodontia: the ad - vantage of rigid lms of various sizes in radio - graphing the mouth. Radiology 1925;5(4): 346–348. 16. Pancoast KH, Fay T. Encephalography as the roentgenologist should understand it: an attempt to standardize the procedure. Radi - ology 1930;15(2):173–210. 17. Bernstein MA, Arens RA. Diagnostic ination of the knee joint: a clinical- radiological study. Radiology 1926;7(6): 500–506. 18. Stein IF, Arens RA. Visualization of the pelvic viscera iodized oil and pneumoperi - toneum combined in gynecology. Radiology 1930;15(1):85–92. 19. Hirsch JS. Urography by uroselectan. Radi - ology 1930;15(4):480–486. 20. Friedman LJ. Iodized oil in roentgenology. Radiology 1929;12(2):114–126. 21. Titterington PF. Portable radiographic au - topsy examination. Radiology 1930;15(1): 135–136. 22. Brown R. Congenital deformity of trachea and esophagus: report of case. Radiology 1926;7(2):166a. 23. Bartlett G. X-ray consultation of the future. Radiology 1925;5(5):439–440. 24. Mills HP, Watkins WV. Localization of for - eign bodies in or about the eye. Radiology 1927;8(4):336–357. 25. Brown B. Differential diagnosis of abdomi - nal tumors by the roentgenological method. Radiology 1928;10(1):48–56. SPECIAL REVIEW: Radiology’s Legacy Levine and Kressel Radiology: Volume 274: Number 1—January 2015  radiology.rsna.org 27 Figure 22 Figure 22: Left : Photograph of original radiograph obtained in New York, NY. Right: Photograph of the same image printed from an unretouched negative, as received by wire in Chicago, Ill. This ushered in the future of

6 medical consultations. Details were not
medical consultations. Details were not lost in the procedure, and an accurate analysis of the image was possible. The picture received in Chicago was immediately sent to New York by airplane mail and was published on the following morning by one of the New York newspapers (g 1 from reference 23). Figure 23 Figure 23: Examination of the right eye showed a perforating wound of the sclera on the nasal side about 1.5 mm from the limbus. Radiograph shows a large foreign body 9 × 2 × 4 mm in the vitreous chamber, with one extremity close to the sclera of the temporal side and localization of this fragment so that scleral incision could be made over one end of it, thereby enabling its removal without further trauma” (g 5 from reference 24). Figure 24 Figure 24: (a) Radiograph shows a normal stomach and duodenum in normal position (g 18 from reference 24). (b) Radiograph shows the right lateral aspect of the stomach. The pylorus and duodenum are displaced forward. The interpretation was that of a retroperitoneal tumor, apparently arising from the pancreas (g 19 from reference 25). SPECIAL REVIEW: Radiology’s Legacy Levine and Kressel 26 radiology.rsna.org  Radiology: Volume 274: Number 1—January 2015 Figure 18 Figure 18: (a, b) Note the variation in appear - ance of the urogram at the second ( a ) and third ( b ) examinations. In b , the ureters are in systole and peristalsis (fig 2 from reference 19). Figure 19 Figure 19: Injected maxillary antrum reveals hypertrophic mucosa (g 3 from reference 20). Figure 20 Figure 20: Radiograph of the right leg shows fracture. Exposure was made through the casket. The article states, “During the St. Louis cyclone of September, 1927, a workman for one of the large factories was killed. Eighteen months later, the exact cause of death was a disputed question. An autopsy committee was appointed to determine whether death had been caused by direct violence or by burns from bursting steam pipes. The con - dition of the body at the time of death prevented thorough embalming. Upon exhumation, it was found that the poor state of preservation prevented the body’s removal from the casket.… The only pathology from a radiographic standpoint was a compound fracture through the middle of the right tibia and bula. This examination was conducted in the eldhouse of the cemetery, with the thermom - eter unfortunately standing close to zero” (g 1 from reference 21). agents were administered, and showed both normal and abnormal ndings. We radiologists take many of these concepts for granted today in the era of cross-sectional imaging. We often think of early medical ra - diography being performed with large cumbersome devices that were available in only select locations within medical Figure 21 Figure 21: Radiograph shows the retention of two drams of water and barium mixture, with the lower level of the meal being opposite the fourth dorsal vertebra. A diagnosis of complete occlusion of the esophagus was made. The following day, the infant died. An autopsy showed that the barium mixture had reached the stomach; however, the pouch as shown on the radiograph did not have an outlet. On passing a soft probe from the stomach upwards, it was found that there was an opening into the trachea at the level of the fth dorsal vertebra and that the barium had reached the stomach by passing down through the vocal cords and the upper part of the trachea through a congenitally formed passage into the lower part of the esophagus and then into the stomach (g 1 from reference 22). facilities. However, articles in the 1920s showed use of portable x-ray machines in the diagnositic process (Fig 20 ) (21) and as a forensic tool in the de - termination of cause of death (Fig 21 ) (22). Remote diagnosis, an advance popularized in the late 1990s, was de - scribed in 1925 in a short two-page description of transmission of soft- copy radiographs (Fig 22) (23). In the aforementioned article, Guy Bartlett of the General Electric Company de - scribed a scenario where “an x-ray lm is made of the injury, sent by wire immediately to specialist in a distant city, and within an hour an answer is received by telegraph.” The article SPECIAL REVIEW: Radiology’s Legacy Levine and Kressel Radiology: Volume 274: Number 1—January 2015  radiology.rsna.org 25 visualized or, equally importantly, to assess function. One can only imagine the excitement felt by early radiologists Figure 16 Figure 16: Radiograph of a normal knee joint. The suprapatellar pouch (A) normally assumes the shape of a gall bladder. The posterior compartment of the knee joint (B) is elliptical in shape. The in - frapatellar space (C) is usually triangular in shape, with the apex at the patella and the base slightly irregular due to the position of the cartilage. Spaces A , B , and C were lled with carbon dioxide gas (g 1 from reference 17). Figure 17 Figure 17: (a) Radiograph shows the right hydrosalpinx contains a “pearl cluster of lipiodol.” (b) Explana- tory diagra

7 m. The article states, “With our increas
m. The article states, “With our increased experience in the use of combined iodized oil and pneu- moperitoneum in gynecology, we feel that we may recommend this method as a safe and precise means of pelvic diagnosis” (g 1 from reference 18). Figure 15 Figure 15: (a) Anteroposterior image in a patient with epilepsy and marked mental deciency shows bi - lateral atrophy in the island of Reil areas, poor lling of the cortical air spaces, and only partial drainage of the right ventricle (g 7 from reference 16). (b) Photograph shows the position of the patient to obtain an anteroposterior ventriculogram with the Bucky diaphragm tilted forward, the head of the patient xed, and the plate changer and the body in position for this exposure. Note that the height of the chair permits the position of the patient to be changed without disturbing the height of the Bucky diaphragm. The Bucky magnetic shift is operated from the transformer control (g 2 from reference 16). and that physiologic function could be evaluated. Many different types of contrast, ranging from negative con - trast from air or gas to positive con - trast with iodine, and methods of con - trast application were described (Figs 15–19 ) (16–20). As stated in an article published in 1929, “Startling revela - tions have been made in almost every branch of medicine and surgery in the past decade, but few have attracted the attention of the medical profession as has the use of a contrast medium in roentgen diagnosis….the iodized oils have almost unlimited uses” (20). Furthermore, an article published in 1930 stated, “Intravenous urography will undoubtedly solve many urological problems without the aid of cystoscopy and ureteral catheterization, and will probably replace catheter pyelography as the method of choice” (19). These articles described the thought process behind the use of contrast agents, de - scribed the manner in which contrast when they realized that the urinary tract and gastrointestinal tract could in fact be visualized radiographically SPECIAL REVIEW: Radiology’s Legacy Levine and Kressel 24 radiology.rsna.org  Radiology: Volume 274: Number 1—January 2015 Figure 12 Figure 13 Figure 13: The relative size of the fetal head and the mother’s pelvis is better seen in the anteroposterior view. The article states: “The amount of radiation required to make two or three exposures of the abdomen is well below the minimum dosage which would have any injurious biological effect on mother or fetus. As a matter of fact, the x-ray department of the Magee Hospital has made all sorts of roentgen examinations on pregnant women daily for the last ten years, and we have yet to see the rst case in which there is any suspicion that such examination has done harm” (g 1 from reference 14). use of x-rays in imaging and therapy, there was a noticeable lack of under - standing of the dangers of radiation used for diagnostic imaging. For ex - ample, in descriptions of obstetric imaging, there was recognition of po - tential harm from radiation exposure; however, this tended to be underem - phasized, since it was believed that harm would be minimal if only a few images were obtained. Specically, in the article from which Figure 13 was taken, Grier states, “The amount of radiation required to make two or three exposures of the abdomen is well below the minimum dosage which would have any injurious biological ef - fect on mother or fetus. As a matter of fact, the X-ray department of the Magee Hospital has made all sorts of roentgen examinations on pregnant women daily for the last ten years, and we have yet to see the rst case in which there is any suspicion that such examination has done harm” (14). Such a quote clearly demonstrates the importance of population studies and statistical analysis rather than the an - ecdotal reporting of experience. Descriptions of a technique, such as how to obtain dental x-rays, pub - lished in 1925, reveal a remarkable similarity to techniques used today (Fig 14 ) (15). The imaging apparatus has changed over the years, a move - ment from analog to digital images has occurred, and overall radiation dose has decreased; however, given the many improvements in hardware and software and a better understanding of exposure and voltage, the key informa - tion provided by the images remains similar to the information provided al - most a century ago. A noteworthy early development was the use of contrast agents to im - prove the scope of anatomic detail Figure 12: (a) Diagram of a working Circuit crest meter developed by Dr Rieber to improve methods of measuring electrical input to x-ray tubes (g 1 from reference 13). (b) Photograph of a dosimeter (g 3 from reference 13). Figure 14 Figure 14: Dental radiograph (g 4 from reference 15). SPECIAL REVIEW: Radiology’s Legacy Levine and Kressel Radiology: Volume 274: Number 1—January 2015  radiology.rsna.org 23 the time. It was only over time that practitioners and journals adopted a more standardized approach toward presentation

8 of images. Structured quality assurance
of images. Structured quality assurance and peer review were not yet a part of im - aging departments; however, in these early articles, authors commonly shared their personal experiences with pitfalls in diagnoses. Remarkably, the poten - tial harms of overdiagnosis or an overly pessimistic prognosis based on imaging ndings were emphasized (Fig 8) (9). Although Radiology has always been a clinically oriented journal, im - portant and classic observations on the behavior of x-rays were published in the journal (Fig 11) (12). For ex - ample, the article by Compton (12), described his work that eventually led to a Nobel prize. Safety issues were of concern, with attempts be - ing made to standardize the radiation dose. There were many articles on radiation physics, use of Bucky dia - phragms, and spot lms. Figure 12 (13) shows some of these aspects of early publications. While there was interest in devel - oping a better understanding of the Figure 9 Figure 9: Radiograph shows chronic miliary tuberculosis. Note the left-right inversion of the image, as standardized projection orientation was not yet estab - lished (g 1 from reference 10). Figure 10 Figure 10: Radiograph shows bone changes due to absorption without amputation. The most notable changes occur in the phalanges of both extremities and in the nasal septum, whereas the bones of the trunk and the long bones of the limbs are unaffected. Like other symptoms of leprosy, there is often a curious limitation to the most peripheral parts. The time factor involved in the loss of bone is widely variable, but most often the process is very slow, with 10–30 years passing while the digits are gradually resorbed (g 2 from reference 11). diagnosis and treatment of this disease. What is a bit surprising when viewing these older images is the lack of stan - dardization of the black on white (Fig 8) (9) and even the left-right (Fig 9) (10) view of the images, with different viewing techniques being used even within individual articles. No doubt this large degree of variation in techniques characterized the state of practice at Figure 11 Figure 11: When an x-ray quantum is scattered by an elec- tron at an angle  , the electron recoils at an angle  , thereby removing some energy from the quantum and reducing its frequency (from 3 from reference 12). SPECIAL REVIEW: Radiology’s Legacy Levine and Kressel 22 radiology.rsna.org  Radiology: Volume 274: Number 1—January 2015 Figure 7 Figure 7: (a) Radiograph shows an open safety pin in the upper third of the esophagus with the point down (g 1 from reference 8). (b) Photograph of the patient 2 hours after removal of the safety pin (g 2 from reference 8). contributions read like review articles until the end of the report, when the new information and the authors’ per - sonal experience and views are shared with the reader. Correlation with gross appearance, function, and pathologic ndings were common. Caffey’s classic description of lead lines in 1931 (Fig 5 ) is one such example (5). Commentary on articles was also frequently incor - porated into the text. For example, the importance of a personal interaction between experts in pathology and ra - diology was encouraged. Dr Robertson wrote, “The pathologist is usually more than willing to cooperate. Will not the roentgenologist fall short of his duty and privilege if he does not seize these opportunities? We, as physicians, owe it to the living that our deaths shall not be total losses” (6). Original articles were the spring point for many of the named abnor - malities that we use in general prac - tice today. For example, the 1940 ar - ticle by Dr Hill and Dr Sachs entitled The Grooved Defect of the Humeral Head: A Frequently Unrecognized Complication of Dislocations of the Shoulder Joint was the rst published description of what we now term the Hill-Sachs deformity (Fig 6) (7). The focus of these articles was typ - ically on the patient (Fig 7) (8), with - out sophisticated statistical analysis. It is evident from perusing these articles that the questions being answered were pertinent to patient care of the times (Figs 8–10 ) (9–11). A common disease of the day, tuberculosis, was featured frequently, with commentary on both Figure 8 Figure 8: Radiograph shows a cavity in a patient with tuberculosis. Although there has been considerable destruction in this patient, the tendency to form brous tissue, as evidenced by the rm wall to the cavity, the bands of opaque shading in each upper lobe, and calci - cation in each hilum lead one to conclude that repair is occurring far more rapidly than destruction. Thus, a favorable prognosis was expressed (g 3 from reference 9). Figure 6 Figure 6: After correction of the third dislocation, a large groove is visible when the humerus is in marked internal rotation (g 3 from reference 7). SPECIAL REVIEW: Radiology’s Legacy Levine and Kressel Radiology: Volume 274: Number 1—January 2015  radiology.rsna.org 21 Figure 3 Figure 3: (a) Radiograph

9 shows stones grouped in the urinary blad
shows stones grouped in the urinary bladder in an almost spherical mass measuring 65  60 mm (g 1 from reference 3). (b) A total of 25 faceted stones were removed (g 2 from reference 3). Figure 2 Figure 2: Chondroma of the left bula (gs 2 and 3 from reference 2). These early articles are based on conventional radiography and provide the foundation for both the way we think about extracting diagnostically important information from images and the spectacular technologic achieve - ments in the decades that followed. Thus, we decided to include in this issue of Radiology a pictorial essay of select gures and our thoughts on the articles themselves to give a taste of the beginnings of so many of our current imaging and research activities. The rst salient observation when reading the early works in Radiology is that the early articles were predomi - nately descriptive (Figs 1–3 ) (1–3) and relied heavily on personal experience. There was commonly no descrip - tion of the study methods, no control groups, and very little, if any, quantita - tive analysis. Early radiologists shared their observations obtained through careful scrutiny of images (Fig 4 ) (4) and comparison of their observa - tions with those reported previously by others. In fact, many of these early Figure 4 Figure 4: Diagram drawn over a normal heart in a child aged 9 years obtained at a distance of 6 feet. DA = descending aorta, GV = great vessels, LA = left auricle, LV = left ventricle, ML = middle left diameter, MR = middle right diameter, PA = pulmonary artery, RA =, right auricle; RV = right ventricle. Middle right diameter (3.5 cm) plus middle left diameter (7 cm) equals 10.5 cm, which is half the diameter of the chest (g 1 from reference 4). Figure 5 Figure 5: Radiograph in an 18-month-old child with lead poisoning. Note the bands of increased opacity at the distal ends of the radii, ulna, and metacarpals (g 1 from reference 5). SPECIAL REVIEW: Radiology’s Legacy Levine and Kressel 20 radiology.rsna.org  Radiology: Volume 274: Number 1—January 2015 REVIEWS AND COMMENTA R Y  SPECIAL REVIEW Published online 10.1148/radiol.14142265 Content code: Radiology 2015; 274:20–28 1 From the Radiology Editorial Ofce, 800 Boylston St, 15th Floor, Boston, MA 02199. Received September 23, 2014; nal version accepted September 23. Address corre - spondence to H.Y.K. (e-mail: hkressel@rsna.org ).  RSNA, 2015 Deborah Levine, MD Herbert Y. Kressel, MD Radiology’s Legacy: Rummaging through the Attic 1 T ucked away in hallways and cor - ners of the Radiology Editorial Offices in Boston sit all of the bound volumes of Radiology since its rst publication in 1923. These sit gathering dust, with little regular use. Once in a great while, we would go to the shelves to retrieve an article re - lating to a more current manuscript on which we were working. With the introduction of the Radiology Legacy Collection, all Radiology articles dat - ing back to the rst issue in 1923 be - came searchable and readily available online. Radiological Society of North America members have access to this archive and can see the articles as originally published (in Portable Document Format [or PDF]). In spite of the ease of search and access, we rarely had occasion to draw on this collection in our daily work. Hap - pily, the RSNA Centennial celebra - tion provided a great opportunity for the Editors and other participants in the “Golden Oldies” effort to search through the collection and reconnect with the history of the development of clinical imaging and imaging science though the pages of the journal. This task may be likened to rummaging through one’s attic as a child to learn about great grandparents, long gone, whose sepia-toned photographs allow one to see these forbearers as people, rather than as just names. In even a casual encounter with this collection, one is stuck by the magni - tude of change in the way imaging is performed and the way images are ana - lyzed, as well as by the vision, intellect, and imagination of the earlier investi - gators. One wonders what these early investigators might have accomplished if they had the imaging tools available to them that we now use on a daily basis. Early on in the development of the “Golden Oldies” lists, we realized that with the approach we were taking, many interesting and visually striking images would not be included, and moreover, the radiograph and its im - portance might not be fully captured. However, radiography forms the basis of the majority of our radiology history. Figure 1 Figure 1: (a) Circumscribed defect in the pars- media, with striations in the middle of the defect due to the interposition of the barium mixture between the villous protrusions. (b) Gross pathologic appear - ance (g 1 from reference 1). Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarig