RESUMO
BACKGROUND: The conventional approach to treatment for Paget's disease of the breast has been mastectomy, but there is an increasing trend to consider breast-conserving surgery (BCS) followed by radiotherapy (RT) in these patients. This study aimed to provide an updated systematic review and meta-analysis comparing outcomes after BCS with RT versus mastectomy in the treatment of Paget's disease of the breast. METHODS: Studies before May 2021 were included. Primary outcomes were overall survival and local recurrence. Separate analyses of Paget's disease associated with ductal carcinoma in situ (DCIS) and invasive ductal carcinoma (IDC) were undertaken. Meta-regression was used to adjust for imbalance in the proportion of IDC among patients selected to undergo BCS versus mastectomy. RESULTS: Overall survival in patients with Paget's disease who underwent BCS with RT was higher than for those who underwent mastectomy with pooled mortality hazard ratio (HR) of 0.68, (95% per cent c.i. 0.45 to 1.01). Patients with Paget's disease with DCIS had higher overall survival after BCS with or without RT versus mastectomy, with adjusted HR of 0.14 (0.10 to 0.20) and 0.28 (0.22 to 0.36), respectively. For patients with Paget's disease and IDC, overall survival was lower for BCS with or without RT versus mastectomy, with adjusted HR of 0.84 (0.57 to 1.25) and 1.64 (1.04 to 2.58), respectively. In Paget's disease and IDC, local recurrence risk was much higher for BCS with RT, RR 26.8 (1.60 to 456) versus without RT, RR 51.8 (6.80 to 391). In patients with Paget's disease and DCIS, risk of local recurrence versus mastectomy was lower for BCS with RT 0.72 (0.11 to 4.50) but slightly higher for BCS alone 1.38 (0.09 to 21.20). CONCLUSION: BCS with RT may be a comparable treatment alternative to mastectomy for patients with Paget's disease with DCIS, and for selected patients with Paget's disease and IDC.
RESUMO
The tremors of Parkinson's disease (PD) and essential tremor (ET) are known to have overlapping characteristics that make it complicated for clinicians to distinguish them. While deep learning is robust in detecting features unnoticeable to humans, an opaque trained model is impractical in clinical scenarios as coincidental correlations in the training data may be used by the model to make classifications, which may result in misdiagnosis. This work aims to overcome the aforementioned challenge of deep learning models by introducing a multilayer BiLSTM network with explainable AI (XAI) that can better explain tremulous characteristics and quantify the respective discovered important regions in tremor differentiation. The proposed network classifies PD, ET, and normal tremors during drinking actions and derives the contribution from tremor characteristics, (i.e., time, frequency, amplitude, and actions) utilized in the classification task. The analysis shows that the XAI-BiLSTM marks the regions with high tremor amplitude as important in classification, which is verified by a high correlation between relevance distribution and tremor displacement amplitude. The XAI-BiLSTM discovered that the transition phases from arm resting to lifting (during the drinking cycle) is the most important action to classify tremors. Additionally, the XAI-BiLSTM reveals frequency ranges that only contribute to the classification of one tremor class, which may be the potential distinctive feature to overcome the overlapping frequencies problem. By revealing critical timing and frequency patterns unique to PD and ET tremors, this proposed XAI-BiLSTM model enables clinicians to make more informed classifications, potentially reducing misclassification rates and improving treatment outcomes.
RESUMO
The distinction between Parkinson's disease (PD) and essential tremor (ET) tremors is subtle, posing challenges in differentiation. To accurately classify the PD and ET, BiLSTM-based recurrent neural networks are employed to classify between normal patients (N), PD patients, and ET patients using accelerometry data on their lower arm (L), hand (H), and upper arm (U) as inputs. The trained recurrent neural network (RNN) has reached 80% accuracy. The neural network is analyzed using layer-wise relevance propagation (LRP) to understand the internal workings of the neural network. A novel explainable AI method, called LRP-based approximate linear weights (ALW), is introduced to identify the similarities in relevance when assigning the class scores in the neural network. The ALW functions as a 2D kernel that linearly transforms the input data directly into the class scores, which significantly reduces the complexity of analyzing the neural network. This new classification method reconstructs the neural network's original function, achieving a 73% PD and ET tremor classification accuracy. By analyzing the ALWs, the correlation between each input and the class can also be determined. Then, the differentiating features can be subsequently identified. Since the input is preprocessed using short-time Fourier transform (STFT), the differences between the magnitude of tremor frequencies ranging from 3 to 30 Hz in the mean N, PD, and ET subjects are successfully identified. Aside from matching the current medical knowledge on frequency content in the tremors, the differentiating features also provide insights about frequency contents in the tremors in other frequency bands and body parts.
Assuntos
Tremor Essencial , Doença de Parkinson , Humanos , Tremor , Inteligência Artificial , Redes Neurais de Computação , Peso ao NascerRESUMO
BACKGROUND: Percutaneous transluminal angioplasty (PTA) is widely used as a first-line revascularisation option in patients with chronic limb threatening ischemia (CLTI). This study aimed to evaluate the short-term endovascular revascularisation treatment outcomes of a cohort of Rutherford 6 (R6) CLTI patients, from a multi-ethnic Asian population in Singapore. Patients with R6 CLTI who underwent endovascular revascularisation from June 2019 to February 2020 at Singapore General Hospital, a tertiary vascular centre in Singapore, were included and followed up for one year. Primary outcome measures included number and type of reinterventions required, 3-, 6- and 12-month mortality, 6- and 12-month amputation free survival (AFS), wound healing success and changes in Rutherford staging after 3, 6 and 12 months. RESULTS: Two hundred fifty-five procedures were performed on 86 patients, of whom 78 (90.7%) were diabetics, 54 (62.8%) had coronary artery disease (CAD) and 54 (62.8%) had chronic kidney disease (CKD). 42 patients (48.8%) required reintervention within 6 months. Multivariate analysis revealed that the presence of CAD was a significant independent predictor for reintervention. Mortality was 15.1%, 20.9% and 33.7% at 3, 6 and 12 months respectively. AFS was 64.0% and 49.4% at 6 and 12 months. Inability to ambulate, congestive heart failure (CHF), dysrhythmia and CKD were significant independent predictors of lower 12-month AFS. CONCLUSIONS: PTA for R6 CLTI patients was associated with relatively high mortality and reintervention rates at one year. CAD was an independent predictor of reintervention. More research is required to help risk stratify which CLTI patients would benefit from an endovascular-first approach versus conservative treatment or an immediate major lower extremity amputation policy.