Thermology in Breast Oncology
Quantitative Digital Radiometric Telethermology in Breast Oncology
By Philip P. Hoekstra, III
Introduction: Diagnostic medical infrared imagery has languished since its introduction in the 1960's for a number of reasons. First, it was not evolved from any existing practice or technology within Medicine nor was it requested by Medicine. Rather it was presented to a reluctant Medicine by as a high-technology benefit of the military intelligence development of modern infrared sensors. Second, the companies that have produced the thermographs had no other established medical products and marketed them to Medicine only as a side venture to their military and engineering products. Third, though diagnostic medical infrared imaging integrates with classic principles of Medicine, it had no existing empirical or basic science basis in medicine on which to infer meaningful diagnoses. Fourth, infrared imaging did not easily merge with any existing medical discipline. It is a medical imaging technology to be sure and hence would fall into the domain of Radiology. However, as the basis for medical infrared imaging is functional rather than structural, it is completely different from all other aspects of Radiology. This has presented a conceptual gap across which radiologists have not been motivated to bridge. Fifth, a major effort to develop medical infrared imaging in the 1970's came in the form of the National Cancer Institute's Breast Cancer Detection and Demonstration project. That project was ill conceived, poorly executed and had results misrepresented as a failure of infrared imaging in medicine rather than a failure of the project.
Background: Medical infrared imaging is surely based in physiology; as such its fundamentals are not as amenable to study, as are the anatomically based imaging modalities. Despite formidable obstacles, medical infrared imaging has been developed by stubborn investigators and greatly aided by basic scientists to now stand as a practical technology with applications in neurology, vascular medicine and breast oncology. This development originated in multi-disciplinary centers at the hands of physicians such as Harold Issard, a radiologist from Temple University, that developed empirical rationales based on their extensive understanding of the applied patho-physiology and a vast clinical experience integrating the thermal images and their patient's condition. The first meaningful large-scale study of infrared imaging and breast oncology took place at the Cancer Institute of Pasteur University in Marseilles France and was published in 1975. This retrospective study took place in a comprehensive multi-disciplinary center and incorporated the experience of medical infrared imaging from tens of thousands case studies over many years. The study yielded the first objective criteria for the analysis of thermal images for breast oncology. This study determined the importance of high-emission vascular-like features as a diagnostic characteristic of infrared imaging for breast oncology. Another significant development of basic science related to breast oncology would also occur in 1975 with a publication of Judah Folkman's theory of neo-angiogenesis of solid malignant tumors. Folkman's theory offered a rational for the important association of atypical vascular-like features in infrared breast oncology. These neo-angiogenic blood vessels must be developed early in the life of a solid malignant tumor, certainly by three (3) mm in diameter, the practical limit of diffusion and into dense convoluted networks. During the late 1980's a group of anatomist from the University of Essen in Germany demonstrate the primitive lacunae structure of the neo-angiogenic vessels and their lack of a regular endothelial layer and complete lack of vascular smooth muscle. This meant neo-angiogenic vessels could not respond to the physiologic modulation by the autonomic nerve system, vasodilators or constrictors. The thermal energy manifest in infrared imagery of the body is a direct manifestation of the inherent inefficiencies of metabolism and the effect of circulating blood. Certainly, infrared energy is emitted superficially and certainly is strongly affected by skin perfusion but the characteristic high-energy vascular-like features associated with breast oncology are not frequently superficial. The question arises as to the mechanism responsible for the infrared vascular-like 'signatures' associated with breast oncology. An answer to this question came from the ability of infrared imagers in planetary orbit to assess subterranean features. Plainly put, if there is a suitable gradient to dissipate the superficial thermal energy, such as cool ambient air; high-energy thermal features from deep will 'float' to the surface by conduction. A higher gradient will facilitate the infrared 'signatures' of deeper high thermal energy especially through a relatively homogenous media. In medicine this effect is limited by ambient air sufficiently cold to causing shivering and the paradoxical hunting phenomenon of Lewis. Further, non-homogenous tissue will dissipate emerging thermal energy and obscure its detail. Fortuitously, human skin, irrespective of its pigmentation, emits infrared energy in essentially perfect proportion to its temperature and the human female breast is relatively homogenous tissue. Conversely, infrared imagery has found little use in the study of deep visceral organs that may be encapsulated, surrounded by heterogeneous tissues and approximated by large-caliber blood vessels.
To date, the development of thermal imaging technology has been directed to its industrial and engineering applications. Infrared imaging has applications that range from dynamic evaluation of tire stress-related failure to astronomy and its practical value has given impetus to a rapid succession of powerful features. The instruments of the 1950's were as stone knives and bear claws to the instruments available today. The first to the fourth generation thermographs were bulky, slow, cranky complex mechanical-optical analog scanning instruments with poor spatial resolute that recorded to photographic film or videotape. These early thermographs did not lend themselves to a quantitative means of analysis and were often little more than extensions of human vision. However, the fifth generation instruments originated in the mid-1990's and feature a dynamic interfaced with a modern digital computer. These instruments are frequently indium antimony focal plane or standing arrays with approximately seventy-five thousand individual detectors and few or no moving parts. The best of the current generation thermographs are twelve (12) to twenty-four (24) bit digital radiometric live images with an instantaneous field of view of better than sixty (60) micro-radians and a spatial resolution better than twenty 20 milliKelvin (20 mK). These images are recorded by a modern microcomputer (PC) as digital data, preserving the radiometric values of each detector pixel.
Medicine has been a qualitative sanctuary in an increasing quantitative world. Despite the perceptions of voluminous numbers as measures in units of millimeters of mercury or milligrams per deciliter, Medicine rarely integrates this data in any quantitative means of analysis for diagnosis or treatment. Perhaps it is that students with a predilection for number concepts gravitate towards computer science, engineering or mathematics rather than Medicine. It is an irony of our age that the marketing of corn flakes incorporates more quantitative and statistical analysis than the diagnosis or treatment of heart disease. This is not to denigrate the accomplishments of physicians as often they develop remarkable judgments. However, this is frequently due to detailed training and vast experiences that program their personal organic albeit qualitative reasoning devices. Harold Issard was able to achieve a consistent eighty-eight (88%) percent sensitivity in diagnosing early breast malignancy using simple infrared images with only the use of his experienced eye; educated by the integrated results of forty years of clinical experience. Currently in the World there are but a handful of thermologists with sufficient training and vast experience to qualify as 'expert readers' of breast infrared images. A majority of them do not employ any quantitative means of analysis. Instead they rely on pattern recognition alone to judge an abnormal patient study. There is a general agreement among the 'expert readers' on diagnostic indicators from breast infrared images and a reporting method based upon the Marseilles system. Some of them have objective criteria they employ while others rely strictly on subjective albeit experienced judgment.