How do AI markings on screening mammograms correspond to cancer location? An informed review of 270 breast cancer cases in BreastScreen Norway.

OBJECTIVES: To compare the location of AI markings on screening mammograms with cancer location on diagnostic mammograms, and to classify interval cancers with high AI score as false negative, minimal sign, or true negative. METHODS: In a retrospective study from 2022, we compared the performance of an AI system with independent double reading according to cancer detection. We found 93% (880/949) of the screen-detected cancers, and 40% (122/305) of the interval cancers to have the highest AI risk score (AI score of 10). In this study, four breast radiologists reviewed mammograms from 126 randomly selected screen-detected cancers and all 120 interval cancers with an AI score of 10. The location of the AI marking was stated as correct/not correct in craniocaudal and mediolateral oblique view. Interval cancers with an AI score of 10 were classified as false negative, minimal sign significant/non-specific, or true negative. RESULTS: All screen-detected cancers and 78% (93/120) of the interval cancers with an AI score of 10 were correctly located by the AI system. The AI markings matched in both views for 79% (100/126) of the screen-detected cancers and 22% (26/120) of the interval cancers. For interval cancers with an AI score of 10, 11% (13/120) were correctly located and classified as false negative, 10% (12/120) as minimal sign significant, 26% (31/120) as minimal sign non-specific, and 31% (37/120) as true negative. CONCLUSION: AI markings corresponded to cancer location for all screen-detected cancers and 78% of the interval cancers with high AI score, indicating a potential for reducing the number of interval cancers. However, it is uncertain whether interval cancers with subtle findings in only one view are actionable for recall in a true screening setting. CLINICAL RELEVANCE STATEMENT: In this study, AI markings corresponded to the location of the cancer in a high percentage of cases, indicating that the AI system accurately identifies the cancer location in mammograms with a high AI score. KEY POINTS: • All screen-detected and 78% of the interval cancers with high AI risk score (AI score of 10) had AI markings in one or two views corresponding to the location of the cancer on diagnostic images. • Among all 120 interval cancers with an AI score of 10, 21% (25/120) were classified as a false negative or minimal sign significant and had AI markings matching the cancer location, suggesting they may be visible on prior screening. • Most of the correctly located interval cancers matched only in one view, and the majority were classified as either true negative or minimal sign non-specific, indicating low potential for being detected earlier in a real screening setting.

Mammographic morphology and distribution of calcifications in ductal carcinoma in situ diagnosed in organized screening.

BACKGROUND: Ductal carcinoma in situ of the breast (DCIS) represents a challenge in mammographic screening due to its unknown progression into invasive cancer. The majority of the DCIS is detected due to signs of calcifications on the mammograms. PURPOSE: To analyze the combinations of mammographic morphology and distribution of calcifications by Van Nuys nuclear grade (grade). MATERIAL AND METHODS: A total of 217 DCIS diagnosed in women aged 50-69 years old who participated in the Norwegian Breast Cancer Screening Program in the period November 1995 to December 2007 were reviewed by four breast imaging specialists. The mammograms were classified according to the morphology and distribution of the calcifications, using BI-RADS nomenclature. Chi square test was used to compare the groups of morphology and distribution by grade. RESULTS: Calcifications were identified in 93% (202/217) of the cases, 15% (30/202) as grade 1 and 74% (149/202) as grade 3. Fine pleomorphic calcifications were seen in 38% (77/202) of the lesions and fine linear and fine linear branching in 31% (62/202). Sixty-nine percent (53/77) of the fine pleomorphic and 84% (52/62) of the fine linear and fine linear branching calcifications were high grade lesions. Grouped distribution was seen in about half of all the cases (104/202). Among the high grade lesions with fine pleomorphic or fine linear and fine linear branching calcifications, 75% (40/53) and 69% (36/52), respectively, had grouped or segmental distribution. CONCLUSION: DCIS presented overlapping groups of morphology and distribution of calcification by grade, but fine pleomorphic and fine linear and fine linear branching calcifications with grouped and segmental distributions were associated with high grade DCIS. Seeking for further knowledge that allows separation of non-high grade from high grade DCIS has to continue to improve the quality of mammographic screening.

Artifacts on Magnetic Resonance Imaging from Electronic Identification Enablement in Silicone Gel Implants Are Not Negligible.

Breast implants filled with silicone gel are used worldwide for cosmetic reasons, or breast reconstruction following risk-reducing or therapeutic mastectomy. The importance of identifiable implants is undeniable. A recent development has been the labeling of the implants with a radio-frequency device micro responder chip (RFID). We examined a patient with silicone implants containing RFID chips with magnetic resonance imaging and were surprised by the artifacts caused by the RFID chip. We raise the question if the benefits of RFID-labeled silicone implants outweigh the drawbacks of magnetic resonance artifacts caused by the RFID chip itself.

Sensitivity of MRI versus conventional screening in the diagnosis of BRCA-associated breast cancer in a national prospective series.

We wanted to compare the sensitivities of breast magnetic resonance imaging (MRI) and the conventional screening programme consisting of mammography (XRM) +/- ultrasound for early diagnosis of breast cancer in BRCA1/2 mutation carriers. BRCA1/2 mutation carriers were examined prospectively by both breast MRI and XRM +/- ultrasound. Eight hundred and sixty-seven MRI examinations were carried out in 445 BRCA1 and 46 BRCA2 mutation carriers. A total of 25 cancers were observed, five (20%) as interval cancers. At the time of diagnosis, sensitivity to detect cancer was 19/22=86% for MRI and 12/24=50% for XRM. Twenty-one were examined by both methods at the time of diagnosis. In the19 BRCA1 mutation carriers among them, MRI had a sensitivity of 1/3(33%) to diagnose DCIS and 15/16 (94%) among the invasive cancers. For XRM the sensitivities were 1/3(33%) for DCIS, 3/7(42%) for pT1b, 3/6(50%) for pT1c, and 3/3/100%) for pT2. In the two BRCA2 mutation carriers, both were demonstrated by breast MRI, neither was detected by XRM. Breast MRI had increased sensitivity compared to XRM to diagnose all cancers staged less than pT2.

Do radiologists agree on the quality of computed tomography enterography?

This study aimed to assess variation between radiologists evaluating the quality of multi-detector computed tomography enterography. For 40 consecutive examinations, three experienced radiologists independently rated the following quality variables: % length of adequately filled bowel, bowel lumen diameters, bowel wall delineation, superior mesenteric vein, and bowel wall enhancement, artefacts, and total quality. We calculated the mean difference between observers with standard deviation (SD) for continuous variables and % total agreement, exact Fleiss kappa, and P-values (McNemar's test) for categorical variables. Depending on bowel segment (duodenum distal to bulb, jejunum, ileum, terminal ileum), mean difference between observers ranged from two to 33 (SD from 11 to 32) for % length of adequately filled bowel judged subjectively, 0-2 (SD 0-3) mm for smallest bowel lumen diameter and 0-4 (SD 3-7) mm for largest bowel lumen diameter. Agreement on bowel wall delineation was 80%/kappa 0.50 in duodenum, 90%/kappa 0.57 in jejunum, 75%/kappa 0.14 in ileum and 88%/kappa 0.17 in terminal ileum, where ratings differed between observers (P < 0.04). Agreement was 65%/kappa 0.18 for bowel wall enhancement judged subjectively. For contrast enhancement measured in Hounsfield Units, mean difference between observers ranged from two to 11 (SD 12-15) in normal jejunum wall and zero to one (SD 4-5) in the superior mesenteric vein depending on observer pair. Agreement was 78%/kappa 0.12 for image artefacts. Rating of total examination quality (good/optimal versus poor/very poor) differed between observers (P < 0.01); agreement was 60%/kappa 0.41. Many subjective evaluations varied between observers. We believe that measurements of bowel lumen diameters and contrast enhancement may be preferable.