Infrared Thermography in Early Breast Cancer Screening

Dr. Bhowmik is actively involved in the domain of medical infrared thermography for early diagnosis of breast cancer and arthritis. Detection of subtle temperature variations and segmentation of the suspicious hotspot region is a very challenging task due to the limitations of infrared thermography images as illustrated in Figure.1 & Figure. 2. Infrared thermograms are poor in contrast and lacks clear images. Segmentation of suspicious hotspot regions from thermograms remained a challenge to the research community.




Figure 1. The enlarged box region shows that there is a smooth transition of intensity value from one region to other for which segmentation of suspicious hot region from breast thermograms is difficult.

Figure 2. The shape of the hotspot region in each thermogram is unique as shown in (a) & (b). Besides, the size of the hotspot regions are also varying with images.




• Overview of Thermal Imaging :


All objects with a temperature above absolute zero emit infrared radiation from their surface. The Stefan-Boltzmann Law defines the relation between radiated energy and temperature by stating that the total radiation emitted by an object is directly proportional to the object’s area, its emissivity and to the fourth power of its absolute temperature. Since the emissivity of the human skin is extremely high measurements of infrared radiation emitted by the skin can be converted directly into accurate temperature values. This makes infrared imaging an ideal procedure to evaluate surface temperatures of the human body for disease diagnosis purposes. Almost 90% of the infrared energy radiated by the human body is in the range of longer wavelengths 6-14 µm, for which LWIR range of infrared is of particular interest of our research. Abnormalities such as malignancies, inflammation, and infection cause localized increases in temperature which appears as hot spots or as asymmetrical patterns in an infrared thermogram. Even though it is nonspecific, infrared thermography can be used as a tool to detect potential problems in a patient’s physiology. Some common applications are preoperative planning of perforators for free flaps, as an adjunct in burn depth analysis, assessment of response to treatment in haemangioma, as diagnostic test for different types of cancer such as melanoma and breast cancer, diagnosis of carpal tunnel and evaluation of inflammation in arthritis.



• Scope:


In literature, there is only one publically available breast thermogram database that lacks ground truth images and for arthritis related inflammation detection, there is no image database. Thus, it is difficult to evaluate the robustness and efficiency of breast abnormality detection, arthritis detection and hotpot segmentation methods. Considering this, our motive is to design breast thermogram database and infrared knee joint databases for designing and evaluating the new methodologies designed for breast abnormality detection and arthritis related inflammation detection. Moreover, despite of the challenges of infrared thermograms, our motive is to design hotspot region segmentation methods.



• Overview of Research in Early Breast Cancer Screening:


Infrared Thermography in Early Breast Cancer Screening
Due to the unavailability of efficient diagnosis modality for women of young age group, only 0% and 1.9% diagnosis were possible under the age group of 20 years and 20-34 years respectively [5]. Moreover, in comparison to the older women of age above 50, breast cancers grow very faster. Even though, X-ray mammography is considered as the gold standard for breast cancer screening, it can detect a tumor that attain a certain size of 1.68 cm [] which takes almost 7 years from the tumor formation. As illustrated in Figure. 3, in contrary to mammography, Infrared thermography can detect the growth of cancer cells prior to tumor formation and it can detect the presence of breast abnormality 8 year earlier than mammography. Moreover, mammography results in high false positive rate resulting in unnecessary biopsies and considerable radiation exposures. Thus, the mission of Biomedical Image Processing Laboratory is to evaluate the efficiency of Infrared Thermography in Early Breast Abnormality detection and to establish Infrared Thermography as a viable breast screening modality in women of all age group, including the pregnant and nursing women. The laboratory is currently applying infrared thermography in clinical studies that focus on the diagnosis of breast abnormalities. The work of the laboratory is supported by the Department of Biotechnology (DBT), Govt. of India.




Figure. 3. Figure shows the growth of breast cancer cells and the capability of Infrared Thermography to detect the growth of breast tumor in 2nd year.




• Designing of Breast Thermogram Image Dataset :


For designing a ground truth annotated breast thermogram database, a standard breast thermogram acquisition protocol suite to develop a breast thermogram database. Other medical findings like clinical examination report, mammography report, fine needle aspiration cytology reports (if available) are also collected to validate the collected thermograms. Besides, the ground truth images of the suspicious hotspot regions are also annotated with the database. Some sample breast thermogram database along with the ground truth images of the suspicious hot regions are illustrated in Figure 4.




(a)	(b)	(c)	(d)	(e)
Figure. 4 shows (a) Healthy (b) Benign and (c) Malignant Breast Thermograms with (d), (e) ground truth images of (b) benign and (c) malignant thermograms.




• Analysis of Breast Thermograms:


To evaluate the potentiality of infrared thermography in early breast abnormality detection, some research works including: bilateral asymmetry based analysis of breast thermograms, singular value based characterization of abnormality and segmentation of suspicious hotspot regions for breast abnormality prediction have been done. The performance of the developed methods are compared with the state-of-the-art methods.



• Featured Articles:


1. Mrinal Kanti Bhowmik, Usha Rani Gogoi, Gautam Majumdar, Dhritiman Datta, Anjan Kumar Ghosh, Debotosh Bhattacharjee, "Designing of Ground Truth Annotated DBT-TU-JU Breast Thermogram Database towards Early Abnormality Prediction", in IEEE Journal of Biomedical and Health Informatics (J-BHI) (Formerly known as IEEE Transactions on Information Technology in Biomedicine), Vol. 22, 2017, Science Citation Indexed (SCI), Impact factor: 4.217.
2. Usha Rani Gogoi, Mrinal Kanti Bhowmik, Debotosh Bhattacharjee, Anjan Kumar Ghosh, “Singular Value based Characterization and Analysis of Thermal Patches for Early Breast Abnormality Detection", in Australasian Physical and Engineering Science in Medicine, Springer, Vol. 41 (4), pp. 861-879, 2018, Science Citation Index Expanded (SCIE), Impact factor: 1.161.
3. Usha Rani Gogoi, Gautam Majumdar, Mrinal Kanti Bhowmik, Anjan Kumar Ghosh, “Evaluating the Efficiency of Infrared Breast Thermography for Early Breast Cancer Risk Prediction in Asymptomatic Population”, in Infrared Physics & Technology, Elsevier, Vol. 99, pp. 201-211, 2019, Science Citation Indexed (SCI), Impact Factor: 2.379.