RÉSUMÉ
PURPOSE: To establish a reliable machine learning model to predict malignancy in breast lesions identified by ultrasound (US) and optimize the negative predictive value to minimize unnecessary biopsies. METHODS: We included clinical and ultrasonographic attributes from 1526 breast lesions classified as BI-RADS 3, 4a, 4b, 4c, 5, and 6 that underwent US-guided breast biopsy in four institutions. We selected the most informative attributes to train nine machine learning models, ensemble models and models with tuned threshold to make inferences about the diagnosis of BI-RADS 4a and 4b lesions (validation dataset). We tested the performance of the final model with 403 new suspicious lesions. RESULTS: The most informative attributes were shape, margin, orientation and size of the lesions, the resistance index of the internal vessel, the age of the patient and the presence of a palpable lump. The highest mean negative predictive value (NPV) was achieved with the K-Nearest Neighbors algorithm (97.9%). Making ensembles did not improve the performance. Tuning the threshold did improve the performance of the models and we chose the algorithm XGBoost with the tuned threshold as the final one. The tested performance of the final model was: NPV 98.1%, false negative 1.9%, positive predictive value 77.1%, false positive 22.9%. Applying this final model, we would have missed 2 of the 231 malignant lesions of the test dataset (0.8%). CONCLUSION: Machine learning can help physicians predict malignancy in suspicious breast lesions identified by the US. Our final model would be able to avoid 60.4% of the biopsies in benign lesions missing less than 1% of the cancer cases.
RÉSUMÉ
Objective: The aim of this work was to describe a technique for building low-cost simulators for training in minimally invasive breast procedures guided by ultrasound (US) and stereotactic mammography (MMG), focusing mainly on training medical professionals studying related areas. Materials and Methods: Low-cost phantoms were developed using organic structures that mimic breast tissue, such as chicken breast and eggplant, and materials that simulate breast lesions. A step-by-step description of the preparation and use of these simulators was made, enabling the reproducibility of the technique by the physicians in training themselves. Results: The low-cost phantoms showed a high degree of echogenic and radiological similarity with human breast tissue, allowing adequate training in minimally invasive procedures. Conclusion: It was possible to build low-cost phantoms that allow training in US- and stereotactic MMG-guided minimally invasive breast procedures.
RÉSUMÉ
OBJECTIVE: There is a lack of studies evaluating the COVID-19 pandemic effect on breast cancer detection according to age-group. This study aimed to assess the pandemic impact on the trend of mammograms, breast biopsies, and breast cancer stage at diagnosis according to age-group. METHODS: This was an ecological time series study by inflection point regression model. We used data from women aged between 30 and 49, 50 and 69, and 70 years or more available in an open-access dataset of the Brazilian public healthcare system (2017-2021). We analyzed the trend of the variables in the pre-pandemic and the pandemic effect on the total time series. RESULTS: The decreasing or stationary trend of mammograms in the pre-pandemic has changed to a decreasing trend in the total time series in all age-groups. Before the pandemic, the increasing trend of breast biopsies has changed to stationary in the total time series in all age-groups. The increasing trend of tumors at stages 0 to II in the pre-pandemic has changed to decreasing or stationary in the total time series. Finally, the increasing trend of tumors at stage III or IV remained increasing in the total time series in all age-groups. CONCLUSION: The pandemic has changed the stationary or increasing trend to a decreasing or stationary trend of mammograms, breast biopsies, and tumors at stages 0 to II but has not influenced the increasing trend of tumors at stages III and IV in all age-groups.
Sujet(s)
Tumeurs du sein , COVID-19 , Femelle , Humains , Adulte , Tumeurs du sein/diagnostic , Tumeurs du sein/épidémiologie , Tumeurs du sein/anatomopathologie , Pandémies , Analyse de série chronologique interrompue , COVID-19/épidémiologie , Mammographie , Dépistage de la COVID-19RÉSUMÉ
Breast pseudoaneurysm is an extremely rare complication of interventional breast procedures. Pregnancy and lactation are associated with increased breast vascularization, which may act as a risk factor. We present the case of a 36-year-old woman in the third trimester of a spontaneous twin pregnancy, who presented with a newly-detected BI-RADS 4 mass in her right breast. The patient requested not to defer a biopsy until after the pregnancy, and an ultrasound-guided breast core biopsy was performed. The patient presented bleeding during the procedure, but no hematomas or other vascular lesions were immediately detected. During follow-up, a breast ultrasound revealed an anechoic circumscribed mass and high-velocity blood flow. The color Doppler showed a spiral blood flow with the Yin-Yang sign, together with a communication channel between the sac and feeding artery. A diagnosis of breast pseudoaneurysm was made. The patient was managed conservatively, and breastfeeding continued normally. This case report highlights the importance of color Doppler in the detection of pseudoaneurysms, and the need to consider deferring invasive breast procedures in pregnant women when possible.
RÉSUMÉ
Papillary breast lesions are a heterogeneous group of neoplasms of diverse imagenological, clinical and morphological presentation that display different behaviour, prognosis and, therefore, controversial diagnosis and management. The aim of this study is to propose an algorithm for the management of mammary lesions. MATERIALS AND METHODS: We show a retrospective review of breast imaging reporting, percutaneous needle biopsy information, histological-pathological reports and subsequent management. RESULTS: A total of 7,920 biopsies were reviewed. Only 136 biopsies from 130 patients with papillary lesions met the inclusion criteria. There was a correlation between the pathologic findings from percutaneous biopsy and the final surgical histologic result in patients with surgery recommendation in all but 2 (2.12%) cases in which the surgery results were upgraded to a malignant disease. CONCLUSIONS: The algorithm proposed in this paper for the management of mammary lesions significantly reduces the possibilities of upgrading and favours decision making between follow-ups or surgery in patients with papillary lesions of the breast.
RÉSUMÉ
PURPOSE: Mindfulness-based programs can reduce stress and help practitioners to have positive attitudes in their daily lives. This randomized controlled trial evaluated the impact of brief Mindfulness interventions on quantitative and qualitative stress parameters in patients undergoing imaging-guided breast biopsies. METHODS: Eighty-two women undergoing percutaneous imaging-guided breast biopsy were randomized into two groups: MBI group or standard care group. One week before the biopsy procedure, on the waiting room and during the biopsy procedure, the MBI group was exposed to mindfulness techniques and the standard care group received supportive dialogue from the biopsy team. Participants completed questionnaires measuring depression, anxiety and stress, demographics, and medical history, besides evaluating their pain experience through a visual analogue scale for pain and had their systolic and diastolic blood pressure, initial and final temperate, heart rate, oxygen saturation, and salivary cortisol measured. RESULTS: Participation in the mindfulness intervention group was associated with reduced levels of perceived stress, blood pressure, heart rate, and oxygen saturation compared to participation in the standard care group (P values < 0.05). No difference was observed regarding salivary cortisol levels, peripheral temperature, and pain perception between the two studied groups. CONCLUSION: Results indicate that an extremely brief mindfulness intervention is a feasible intervention, suggesting that Mindfulness-based programs may be beneficial to reduce discomfort in acutely stressful settings.
Sujet(s)
Biopsie/effets indésirables , Tumeurs du sein/diagnostic , Tumeurs du sein/psychologie , Pleine conscience/méthodes , Adulte , Anxiété/physiopathologie , Anxiété/thérapie , Biopsie/psychologie , Région mammaire/imagerie diagnostique , Région mammaire/physiopathologie , Tumeurs du sein/imagerie diagnostique , Dépression/physiopathologie , Dépression/psychologie , Femelle , Humains , Adulte d'âge moyen , Mesure de la douleur , Stress psychologique/physiopathologie , Stress psychologique/psychologie , Enquêtes et questionnaires , Résultat thérapeutiqueRÉSUMÉ
Objetivos Describir la experiencia inicial en el uso de un sistema de biopsia por vacío con guía estereotáxica digital para lesiones mamarias. Material y método Se evaluaron retrospectivamente las biopsias bajo estereotaxia mamográfica con sistema de vacío (bav) realizadas en el Centro de Imagenología Mamaria (imad) de la ciudad de Bahía Blanca desde marzo de 2016 a julio de 2017. En el período analizado, se recibieron 329 pedidos de punción, de los cuales 112 correspondieron a biopsias bajo estereotaxia mamográfica con sistema de vacío y 217 a biopsias bajo ecografía. Se realizaron 99/112 (89%) bav, no pudiéndose completar el procedimiento en 13/112 (11%) de los casos. Resultados Se obtuvieron los siguientes resultados patológicos: 20% (20/99) Maligno, 7% (7/99) Alto riesgo y 73% (72/99) Benigno. Se produjeron complicaciones leves que no requirieron ningún tipo de tratamiento posterior en el 8% de los casos: 7/99 reacción vasovagal y 1/99 dolor que dificultó el procedimiento, aunque en todos los casos se pudo finalizar el mismo. Conclusiones Se diagnosticaron lesiones malignas o de alto riesgo en el 27% de los procedimientos realizados.
Objectives To describe our initial experience with vacuum assisted breast biopsies under stereotactic guidance for non-palpable lesions. Materials and method We retrospectively review the vacuum assisted breast biopsies performed at Centro de Imagenología Mamaria (imad) between march 2016 and july 2017. 329 patients requested a breast biopsy. Of them, 112 were vacuum assisted biopsies under stereotactic guidance and 217 were ultrasound guided. We succesfully completed 99/112 (89%) of vacuum assisted biopsies, not being able to reach the target in the remaining 11%. Results The following pathological results were obtained: 20% (20/99) Malignant, 7% (7/99) High risk and 73% (72/99) Benign lesions. Minor complications, which did not require any further treatment, occurred in 8% of cases: 7/99 vasovagal reactions, 1/99 pain during the procedure. Conclusions Malignant and high-risk lesions were diagnosed in 27% of the procedures performed.
Sujet(s)
Humains , Femelle , Tumeurs du sein , Biopsie , RadiochirurgieRÉSUMÉ
BACKGROUND: Image-guided charcoal injection in suspicious breast lesions for preoperative localization is a procedure that has been increasing over the years because it is safer, faster, and more affordable when compared to needle-wire preoperative localization. To date, no complications have been associated with the method. However, in recent years there have been some reports about charcoal granulomas mimicking malignant lesions in some postoperative patients or in a conservative follow-up. PURPOSE: To report a series of 11 cases which had suspicious imaging findings for malignancy and resulted in charcoal granulomas on histopathological analysis. MATERIAL AND METHODS: A database of 1650 patients that attended our center from January 2007 to June 2018 was reviewed and detected 495 patients who had been previously submitted to ultrasound-guided charcoal marking in a breast lesion. Then, patients whose imaging studies were compatible with new suspicious lesions on mammography, breast ultrasound, and/or magnetic resonance imaging and biopsy of this new lesion indicating charcoal granuloma were selected. RESULTS: From 495 patients who had undergone charcoal localization injections in previous biopsies, we selected 11 who had new lesions with malignant characteristics on imaging studies but histopathological analysis resulted in charcoal granuloma. CONCLUSION: Charcoal granuloma should be considered in patients with previous preoperative injection localization, since the residual charcoal in the breast tissue may form granulomas and mimic malignant lesions on follow-up imaging studies.
RÉSUMÉ
Imaging exams are fundamental tools to characterize palpable lesions and to early detect those not identified on the physical examination. However, the correct interpretation of these findings should be done by correlating them with the probable histological diagnosis of the lesion, and performing the appropriate treatment in a timely manner. The Breast Imaging-Reporting and Data System (BIRADS®) is the system used for this association, which characterizes the findings in mammography, ultrasonography and mammary magnetic resonance images, classifying them according to the probability of malignancy. Objective: To correlate imaging features of breast nodules, evidenced by ultrasonography and classified according to BIRADS®, with the histopathological examination results of material obtained through thick needle biopsy of patients from the mastology ambulatory of the Maternity School Assis Chateaubriand that confirm this predictive value of the imaging examination and the impact on surgical indications. Results: We analyzed 110 patient's medical records that fit the inclusion criteria and found that more than 97% of lesions with low suspicion of malignancy, BIRADS® 4A, presented a benign histopathological result. However, all patients with images of nodules with high suspicion of malignancy had histopathological diagnosis of invasive carcinoma. In conclusion, the ultrasonographic features of breast lesions have high predictive value in the final diagnosis of the lesion, supporting the decision of conduct in adequate time in each situation.
Os exames de imagem são ferramentas fundamentais na caracterização de lesões palpáveis e na detecção precoce daquelas não identificadas ao exame físico. Porém, é necessária a interpretação correta desses achados, correlacionando com o provável diagnóstico histológico da lesão, realizando tratamento adequado e em tempo certo. O Breast Imaging-Reporting and Data System (BIRADS®) é o sistema utilizado para essa relação, caracterizando os achados em imagens de mamografia, ultrassonografia e ressonância magnética mamária e classificando-os de acordo com a probabilidade de malignidade. Objetivo: Correlacionar características imagenológicas de nodulações mamárias, evidenciadas por ultrassonografia e classificadas de acordo com o BIRADS®, com os resultados dos exames histopatológicos de material obtido através de biópsia de agulha grossa de pacientes do ambulatório de mastologia da Maternidade Escola Assis Chateaubriand, confirmando esse valor preditivo do exame de imagem e o impacto nas indicações cirúrgicas. Resultados: Foram analisados 110 prontuários de pacientes que se encaixavam nos critérios de inclusão e constatou-se que mais de 97% das lesões com baixa suspeição de malignidade, BIRADS® 4A, apresentaram resultado benigno no exame histopatológico. Entretanto, todas as pacientes com imagens de nódulos com alta suspeição de malignidade tiveram diagnóstico histopatológico de carcinoma invasor. Pode-se concluir que as características ultrassonográficas das lesões mamárias têm alto valor preditivo no diagnóstico final da lesão, fundamentando a decisão de conduta em tempo adequado em cada situação.
RÉSUMÉ
OBJETIVO: A biópsia vácuo-assistida é a forma percutânea de biópsia de microcalcificações que obtém a menor taxa de subestimação, porém, seu custo é alto, havendo interesse em se conseguir formas mais baratas de biópsia vácuo-assistida. O objetivo deste trabalho foi testar um dispositivo portátil de biópsia vácuo-assistida que apresenta custo menor. MATERIAIS E MÉTODOS: Foram biopsiadas 35 pacientes que apresentavam agrupamentos de microcalcificações BI-RADS® 4 ou 5. Foram testados a representatividade dos fragmentos colhidos, as dificuldades na reintrodução da cânula e o número de ciclos de colheita. RESULTADOS: Houve obtenção de calcificações representativas em todas as pacientes. Não houve discordância anatomorradiológica, dificuldade na reintrodução da cânula ou complicações graves. CONCLUSÃO: Os dados permitem concluir que o sistema apresenta boa eficácia na obtenção das amostras e com relação de custo-benefício favorável em relação a outros sistemas para a biópsia de microcalcificações, achados em concordância com outras publicações da literatura.
OBJECTIVE: Vacuum-assisted biopsy is the percutaneous technique of breast biopsy with the lowest underestimation rate. However, the cost of such procedure is high and currently there is a considerable interest in developing less expensive techniques. The present study was aimed at testing a less expensive device for vacuum-assisted biopsy of breast microcalcifications. MATERIALS AND METHODS: Thirty-five patients with clustered microcalcifications classified as BI-RADS® 4 or 5 were submitted to biopsy. Collected specimen appropriateness, difficulties in the reinsertion of the cannula and number of biopsy passes were evaluated. RESULTS: Successful specimens collection was achieved in all of the patients. Histo-radiological disagreement, difficulties in the cannula reinsertion or severe complications were not observed. CONCLUSION: The authors conclude that the method is effective in terms of specimens appropriateness and cost-benefit ratio as compared with of biopsy techniques for breast microcalcifications. Such findings are compatible with data reported in the literature.