The observational single-center study was performed at Kamla Nehru Memorial Hospital and Regional Cancer Centre, Prayagraj, India. In this study, 96 thermograms of 12 female patients in the age group of 25 to 45 years were screened using a thermal camera. Well-defined breast thermogram acquisition protocols [16][24] were strictly followed for capturing images of female subjects. Before the thermography, the subject’s informed consent was taken and they were informed that the breast thermogram and the associated data will be utilized only for the academic research. It was made clear that the identity of the subjects, associated data, and images will remain strictly confidential.
The research study was not reviewed by any ethical committee since this is a non-invasive, non-toxic method of breast imaging conducted as a preliminary study to consider the feasibility and application of further such studies. The research was strictly based on the recommendations stated in the Declaration of Helsinki. Informed consent of the female subjects was taken and recorded with strict consideration regarding maintaining the confidentiality of the subject’s data and images. Before static thermography, patients were asked to abstain from sunlight, heavy meal, or intense exercises. Images of subjects were taken in a chamber with regulated temperature (which was recorded for reference). Since subjects were mostly from a rural background, they were explained about the importance of timely screening and monitoring of overall and breast health. All the personal details of both malignant and benign female subjects including name, age, sex, blood pressure, weight, menopause status, etc., and details of breast condition (symptoms), if any, and corresponding period, were noted down. Information related to surgeries done before or family history of cancer, if any, was also noted down for comparison.
During the process of taking thermograms, only 1 female person was present along with the oncologists. FLIR E40 thermal camera has been used to acquire breast thermal images, with thermal sensitivity of 0.07 ˚C at 30 ˚C and image resolution of 240 × 320 pixels. The distance of 45.7 cm and 76.2 cm was maintained between the subject and thermal camera, to optimally cover the region of interest, and images were selected based on the patient’s framework. FLIR research IR software has been used for transferring and pre-processing acquired breast thermograms. Visible distinguishing features like deformation or nipple inversion were also recorded for correlation. Except for 2 malignant patients with advanced stages of cancer i.e. subject 10 and subject 11, five different views-frontal, left lateral, left oblique, right lateral, right oblique of all subjects were taken.
Wiener filter [28] is applied for noise reduction before segmentation, by producing linear estimation of the original image with minimum mean square error. MATLAB R2020b software is used for post-processing and feature extraction from breast thermograms. Normalization is also done before feature extraction as it helps in monitoring the progression of disease in clinical studies. All other reports of the subjects such as detailed self-examination, CBE, mammography, ultrasonography, MRI, FNAC (Fine needle aspiration cytology), and biopsy, were recorded for comparison. Clinical details of the subjects whose thermograms are taken, along with tumor location and key thermographic observations are given in Table I.
Table I
Clinical Characteristics Of Subjects Along With Tumor Details And Thermograms Findings
Subject | Age | Breast Density | Meno-pause Status | Diagnosis | Breast Side | Histology | Diagnos-tic test for correlati-on | Lesion and Location and type | Thermo-graphic Findings |
1 | 28 | normal fibro-glandular | pre | M | right | carcinoma | Mammo, Biopsy | right breast at 2 o’clock position (ill-defined dense lesion of size 39 × 21mm) | hot spot (R) |
2 | 25 | - | pre | B | both | fibro-adenoma | USG | many locations in both breasts (Diffusely thickened TDLUs) | no hot spot |
3 | 45 | normal glandular | post | M | left | triple-negative infiltrating ductal carcinoma(BIRADS 5) | Mammo, biopsy | left breast at 3 o’clock position(radio-opaque lesion of size 2.2 × 1.9 cm | hot spot in UOQ (L) |
4 | 30 | - | pre | B | left | fibro-adenoma | CBE | - | hot spot (L) |
5 | 32 | dense breast | pre | B | left | hypertrophied auxiliary tail (BIRADS 3) | MRI | - | Extremely small hot spot (L) |
6 | 45 | dense breast | pre | B | right | fibrocystic, cysts in both breasts (R > L) | Mammo FNAC | Tiny cystic lesions | no hot spot |
7 | 37 | - | pre | B | left | severe Infection | - | - | hot spot (L, R) |
8 | 37 | normal fibro-glandular | pre | B | right | fibro-adenosis | Mammo | - | hot spot(R) |
9 | 29 | glandular | pre | B | right | asymmetric breast parenchyma(BIRADS-2) | MRI | - | no hot spot |
10 | 32 | predominantly Fatty | pre | M | Right | triple-negative carcinoma(Post chemo and surgery)(BIRADS 2) | MRI | ILQ right breast(post-operative scar, oval circumscribed collection of size 4.7 × 2.0cm) | large hot spot (R) |
11 | 40 | heterogeneous Fibro-glandular | pre | M | Left | phylloid Tumor (BIRADS 4) | MRI | All quadrants of the left breast(poorly circumscribed, lobulated lesion of size(5.8 × 6.3 × 6.3cm-AP, TR, CC respectively) | NA |
12 | 25 | dense glandular | pre | M | Left | advance cancer, nodal, liver, lung, and bone metastasis (BIRADS 5) | MRI | Multicentrallobulated heterogeneous enhancing mass lesions (L) (one mass of size 3.6 × 4.9 × 3.7 cm in both outer and inner quadrants) | hot spot in UIQ (L) and small hot spot in UIQ (R ) |
M = malignant, malignant = cells divide without control and spread to other body parts, B = benign, benign remain confined and do not spread to other body parts, BIRADS = breast imaging and reporting data system, BIRADS 2 = 0% likelihood of malignancy, BIRADS 3 = less than 2% likelihood of malignancy, BIRADS 4 = probably malignant with 20 to 35% chances of malignancy mass, BIRADS 5 = highly suspicious of malignancy, UOQ = upper outer quadrant of the breast, Triple-negative = estrogen receptor(ER) negative: progesterone(PR) receptor negative: hormone epidermal growth factor receptor-2(HER-2) negative, ductal carcinoma = cells lining the milk duct in the breast becoming cancerous, mammo = mammography, USG = ultrasonography, CBE = clinical breast examination, MRI = magnetic resonance imaging, FNAC = fine needle aspiration cytology, TDLU = terminal duct lobular units or lobules, AP = anteroposterior view, TR = transverse view, CC = craniocaudal view |
Few representative thermograms of the IIITA –KNMH breast thermographic database of some subjects with benign and malignant conditions are shown in Fig. 1.