Using a Machine Learning Approach to Predict the Need for Elective Revision and Unplanned Surgery after Implant-based Breast Reconstruction.

Two-stage implant-based reconstruction after mastectomy may require secondary revision procedures to treat complications, correct defects, and improve aesthetic outcomes. Patients should be counseled on the possibility of additional procedures during the initial visit, but the likelihood of requiring another procedure is dependent on many patient- and surgeon-specific factors. This study aims to identify patient-specific factors and surgical techniques associated with higher rates of secondary procedures and offer a machine learning model to compute individualized assessments for preoperative counseling. A training set of 209 patients (406 breasts) who underwent two-stage alloplastic reconstruction was created, with 45.57% of breasts (185 of 406) requiring revisional or unplanned surgery. On multivariate analysis, hypertension, no tobacco use, and textured expander use corresponded to lower odds of additional surgery. In contrast, higher initial tissue expander volume, vertical radial incision, and larger nipple-inframammary fold distance conferred higher odds of additional surgery. The neural network model trained on clinically significant variables achieved the highest collective performance metrics, with ROC AUC of 0.74, sensitivity of 84.2, specificity of 63.6, and accuracy of 62.1. The proposed machine learning model trained on a single surgeon's data offers a precise and reliable tool to assess an individual patient's risk of secondary procedures. Machine learning models enable physicians to tailor surgical planning and empower patients to make informed decisions aligned with their lifestyle and preferences. The utilization of this technology is especially applicable to plastic surgery, where outcomes are subject to a variety of patient-specific factors and surgeon practices, including threshold to perform secondary procedures.

Distilling Knowledge From an Ensemble of Vision Transformers for Improved Classification of Breast Ultrasound.

RATIONALE AND OBJECTIVES: To develop a deep learning model for the automated classification of breast ultrasound images as benign or malignant. More specifically, the application of vision transformers, ensemble learning, and knowledge distillation is explored for breast ultrasound classification. MATERIALS AND METHODS: Single view, B-mode ultrasound images were curated from the publicly available Breast Ultrasound Image (BUSI) dataset, which has categorical ground truth labels (benign vs malignant) assigned by radiologists and malignant cases confirmed by biopsy. The performance of vision transformers (ViT) is compared to convolutional neural networks (CNN), followed by a comparison between supervised, self-supervised, and randomly initialized ViT. Subsequently, the ensemble of 10 independently trained ViT, where the ensemble model is the unweighted average of the output of each individual model is compared to the performance of each ViT alone. Finally, we train a single ViT to emulate the ensembled ViT using knowledge distillation. RESULTS: On this dataset that was trained using five-fold cross validation, ViT outperforms CNN, while self-supervised ViT outperform supervised and randomly initialized ViT. The ensemble model achieves an area under the receiver operating characteristics curve (AuROC) and area under the precision recall curve (AuPRC) of 0.977 and 0.965 on the test set, outperforming the average AuROC and AuPRC of the independently trained ViTs (0.958 ± 0.05 and 0.931 ± 0.016). The distilled ViT achieves an AuROC and AuPRC of 0.972 and 0.960. CONCLUSION: Both transfer learning and ensemble learning can each offer increased performance independently and can be sequentially combined to collectively improve the performance of the final model. Furthermore, a single vision transformer can be trained to match the performance of an ensemble of a set of vision transformers using knowledge distillation.

Three-dimensional thermophotonic super-resolution imaging by spatiotemporal diffusion reversal method.

Spatial super-resolution in thermophotonic imaging was achieved using a combination of spatial second-derivative forming, spatial gradient adaptive filtering, and Richardson-Lucy deconvolution in conjunction with the construction of an experimental point spread function. When implemented through enhanced truncation-correlation photothermal coherence tomography (eTC-PCT), it was possible to restore blurred infrared thermophotonic images to their prediffusion optical resolution state. This modality was tested in various biological applications and proved to be capable of imaging fine axial cracks in human teeth, well-patterned anatomical subsurface structures of a mouse brain, and neovascularization in a mouse thigh due to the rapid proliferation of cancer cells. This modality was found to be immune to optical scattering and could reveal the true spatial extent of biological features at subsurface depths that conventional thermal imaging cannot reach because of limitations imposed by the physics of spreading diffusion.

Machine-Learning Prediction of Capsular Contraction after Two-Stage Breast Reconstruction.

Background: Two-stage breast reconstruction is a common technique used to restore preoperative appearance in patients undergoing mastectomy. However, capsular contracture may develop and lead to implant failure and significant morbidity. The objective of this study is to build a machine-learning model that can determine the risk of developing contracture formation after two-stage breast reconstruction. Methods: A total of 209 women (406 samples) were included in the study cohort. Patient characteristics that were readily accessible at the preoperative visit and details pertaining to the surgical approach were used as input data for the machine-learning model. Supervised learning models were assessed using 5-fold cross validation. A neural network model is also evaluated using a 0.8/0.1/0.1 train/validate/test split. Results: Among the subjects, 144 (35.47%) developed capsular contracture. Older age, smaller nipple-inframammary fold distance, retropectoral implant placement, synthetic mesh usage, and postoperative radiation increased the odds of capsular contracture (p < 0.05). The neural network achieved the best performance metrics among the models tested, with a test accuracy of 0.82 and area under receiver operative curve of 0.79. Conclusion: To our knowledge, this is the first study that uses a neural network to predict the development of capsular contraction after two-stage implant-based reconstruction. At the preoperative visit, surgeons may counsel high-risk patients on the potential need for further revisions or guide them toward autologous reconstruction.

"The use of bilateral continuous erector spinae plane blocks for postoperative analgesia after right-sided living donor hepatectomy: A feasibility study".

BACKGROUND: Postoperative pain after living donor hepatectomy is significant. Postoperative coagulopathy may limit the use of epidural analgesia, the gold standard for pain control in abdominal surgery. The erector spinae plane block (ESPB) is a novel regional anesthesia technique that has been shown to provide effective analgesia in abdominal surgery. In this study, we examined the effect of continuous ESPB, administered via catheters, on perioperative opioid requirements after right living donor hepatectomies for liver transplantation. METHODS: We performed a retrospective cohort study in patients undergoing right living donor hepatectomy. Twenty-four patients who received preoperative ESPB were compared to 51 historical controls who did not receive regional anesthesia. The primary endpoint was the total amount of oral morphine equivalents (OMEs) required on the day of surgery and postoperative day (POD) 1. RESULTS: Patients in the ESPB group required a lower total amount of OMEs on the day of surgery and POD 1 [141 (107-188) mg] compared the control group [293 (220-380) mg; P < .001]. CONCLUSIONS: The use of continuous ESPB significantly reduced opioid consumption following right living donor hepatectomy.

Machine learning identifies candidates for drug repurposing in Alzheimer's disease.

Clinical trials of novel therapeutics for Alzheimer's Disease (AD) have consumed a large amount of time and resources with largely negative results. Repurposing drugs already approved by the Food and Drug Administration (FDA) for another indication is a more rapid and less expensive option. We present DRIAD (Drug Repurposing In AD), a machine learning framework that quantifies potential associations between the pathology of AD severity (the Braak stage) and molecular mechanisms as encoded in lists of gene names. DRIAD is applied to lists of genes arising from perturbations in differentiated human neural cell cultures by 80 FDA-approved and clinically tested drugs, producing a ranked list of possible repurposing candidates. Top-scoring drugs are inspected for common trends among their targets. We propose that the DRIAD method can be used to nominate drugs that, after additional validation and identification of relevant pharmacodynamic biomarker(s), could be readily evaluated in a clinical trial.

Computer-based video training is effective in teaching basic surgical skills to novices without faculty involvement using a self-directed, sequential and incremental program.

INTRODUCTION: Computer-based video training (CBVT) of surgical skills overcomes limitations of 1:1 instruction. We hypothesized that a self-directed CBVT program could teach novices by dividing basic surgical skills into sequential, easily-mastered steps. METHODS: We developed a 12 video program teaching basic knot tying and suturing skills introduced in discrete, incremental steps. Students were evaluated pre- and post-course with a self-assessment, a written exam and a skill assessment. RESULTS: Students (n = 221) who completed the course demonstrated significant improvement. Their average pre-course product quality score and assessment of technique using standard Global Rating Scale (GRS) were <0.4 for 6 measured skills (scale 0-5) and increased post-course to ≥3.25 except for the skill tying on tension whose GRS = 2.51. Average speed increased for all skills. Students' self-ratings (scale 1-5) increased from an average of 1.4 ± 0.7 pre-elective to 3.9 ± 0.9 post-elective across all skills (P < 0.01). CONCLUSION: Self-directed, incremental and sequential video training is effective teaching basic surgical skills and may be a model to teach other skills or to play a larger role in remote learning.

Intravenous Fluid Management Practices in Kidney Transplant Patients: A Multicenter Observational Cohort Pilot Study.

BACKGROUND: Limited evidence exists with regard to best practices in fluid management during kidney transplantation, which may directly affect the incidence of DGF. The authors of this study embarked on a collaborative observational multicenter pilot study to evaluate fluid administration practices in different transplant centers, with a focus on the relationship between total administered crystalloid volume and its association with DGF. METHODS: Twenty consecutive kidney transplant patients were included from 9 academic medical centers in the United States. One hundred eighty patients were included in the final cohort and variables were compared between patients with and without DGF. Administered crystalloid volume was the primary variable of interest; however, additional patient and surgical variables were compared between patients with and without DGF. Variation in crystalloid administration was explored between centers by comparing median administered crystalloid volumes per kilogram of body weight. Also, unadjusted and adjusted logistic regression analyses were performed to determine which variables were independently associated with DGF. RESULTS: Multivariable regression modeling demonstrated that cold ischemic time and ephedrine use during surgery were independently associated with DGF. There was no independent association between administered crystalloid volume and DGF. CONCLUSION: In this study of patients having kidney transplantation, we did not find an independent association between administered crystalloid volume and DGF, although there was significant variability in crystalloid administration between centers. Our data suggest that DGF was driven mainly by surgical factors such as cold ischemic time. Ephedrine was also independently associated with DGF, which should be explored in future studies.

Phlpp1 is associated with human intervertebral disc degeneration and its deficiency promotes healing after needle puncture injury in mice.

Back pain is a leading cause of global disability and is strongly associated with intervertebral disc (IVD) degeneration (IDD). Hallmarks of IDD include progressive cell loss and matrix degradation. The Akt signaling pathway regulates cellularity and matrix production in IVDs and its inactivation is known to contribute to a catabolic shift and increased cell loss via apoptosis. The PH domain leucine-rich repeat protein phosphatase (Phlpp1) directly regulates Akt signaling and therefore may play a role in regulating IDD, yet this has not been investigated. The aim of this study was to investigate if Phlpp1 has a role in Akt dysregulation during IDD. In human IVDs, Phlpp1 expression was positively correlated with IDD and the apoptosis marker cleaved Caspase-3, suggesting a key role of Phlpp1 in the progression of IDD. In mice, 3 days after IVD needle puncture injury, Phlpp1 knockout (KO) promoted Akt phosphorylation and cell proliferation, with less apoptosis. At 2 and 8 months after injury, Phlpp1 deficiency also had protective effects on IVD cellularity, matrix production, and collagen structure as measured with histological and immunohistochemical analyses. Specifically, Phlpp1-deletion resulted in enhanced nucleus pulposus matrix production and more chondrocytic cells at 2 months, and increased IVD height, nucleus pulposus cellularity, and extracellular matrix deposition 8 months after injury. In conclusion, Phlpp1 has a role in limiting cell survival and matrix degradation in IDD and research targeting its suppression could identify a potential therapeutic target for IDD.

Real-Time Quantification of Cell Internalization Kinetics by Functionalized Bioluminescent Nanoprobes.

Cells transport mass dynamically, crossing cell membranes to maintain metabolism and systemic homeostasis, through which biomolecules are also delivered to cells for gene editing, cell reprograming, therapy, and other purposes. Quantifying the translocation kinetics is fundamentally and clinically essential, but remains limited by fluorescence-based technologies, which are semi-quantitative and only provide kinetics information at cellular level or in discrete time. Herein, a real-time method of quantifying cell internalization kinetics is reported using functionalized firefly-luciferase nanocapsules as the probe. This quantitative assay will facilitate the rational design of delivery vectors and enable high-throughput screening of peptides and other functional molecules, constituting an effective tool for broad applications, including drug development and cancer therapy.

A Bioinspired Platform for Effective Delivery of Protein Therapeutics to the Central Nervous System.

Central nervous system (CNS) diseases are the leading cause of morbidity and mortality; their treatment, however, remains constrained by the blood-brain barrier (BBB) that impedes the access of most therapeutics to the brain. A CNS delivery platform for protein therapeutics, which is achieved by encapsulating the proteins within nanocapsules that contain choline and acetylcholine analogues, is reported herein. Mediated by nicotinic acetylcholine receptors and choline transporters, such nanocapsules can effectively penetrate the BBB and deliver the therapeutics to the CNS, as demonstrated in mice and non-human primates. This universal platform, in general, enables the delivery of any protein therapeutics of interest to the brain, opening a new avenue for the treatment of CNS diseases.

DbVar and DGVa: public archives for genomic structural variation.

Much has changed in the last two years at DGVa (http://www.ebi.ac.uk/dgva) and dbVar (http://www.ncbi.nlm.nih.gov/dbvar). We are now processing direct submissions rather than only curating data from the literature and our joint study catalog includes data from over 100 studies in 11 organisms. Studies from human dominate with data from control and case populations, tumor samples as well as three large curated studies derived from multiple sources. During the processing of these data, we have made improvements to our data model, submission process and data representation. Additionally, we have made significant improvements in providing access to these data via web and FTP interfaces.

A shed form of LDL receptor-related protein-1 regulates peripheral nerve injury and neuropathic pain in rodents.

Injury to the peripheral nervous system (PNS) initiates a response controlled by multiple extracellular mediators, many of which contribute to the development of neuropathic pain. Schwann cells in an injured nerve demonstrate increased expression of LDL receptor-related protein-1 (LRP1), an endocytic receptor for diverse ligands and a cell survival factor. Here we report that a fragment of LRP1, in which a soluble or shed form of LRP1 with an intact alpha-chain (sLRP-alpha), was shed by Schwann cells in vitro and in the PNS after injury. Injection of purified sLRP-alpha into mouse sciatic nerves prior to chronic constriction injury (CCI) inhibited p38 MAPK activation (P-p38) and decreased expression of TNF-alpha and IL-1beta locally. sLRP-alpha also inhibited CCI-induced spontaneous neuropathic pain and decreased inflammatory cytokine expression in the spinal dorsal horn, where neuropathic pain processing occurs. In cultures of Schwann cells, astrocytes, and microglia, sLRP-alpha inhibited TNF-alpha-induced activation of p38 MAPK and ERK/MAPK. The activity of sLRP-alpha did not involve TNF-alpha binding, but rather glial cell preconditioning, so that the subsequent response to TNF-alpha was inhibited. Our results show that sLRP-alpha is biologically active and may attenuate neuropathic pain. In the PNS, the function of LRP1 may reflect the integrated activities of the membrane-anchored and shed forms of LRP1.

Employment of on-line FT-IR spectroscopy to monitor the deprotection of a 9-fluorenylmethyl protected carboxylic acid peptide conjugate of doxorubicin.

A method for accurately determining the end-point, >98% conversion, of the deprotection reaction of a highly toxic 9-fluorenylmethyl (Fm) ester 1b to its corresponding carboxylate 1d in real time by FT-IR spectroscopy is reported. Advantages of this method over analysis by conventional chromatographic means include real time determination of the end-point of a reaction that is time sensitive to by-product formation, and elimination of sampling a highly toxic reaction mixture. The FT-IR method is based on monitoring, in real time, the disappearance of the Fm ester carbonyl band for 1b at 1737 cm(-1), during deprotection by piperidine, and calibration models were established by Partial Least Squares (PLS) regression analysis with high performance liquid chromatography (HPLC) as reference. The best calibration model was built with 5 PLS factors in the spectral range of 1780-1730 and 1551-1441 cm(-1) and resulted in a standard error of cross validation (SECV) of 0.63 mM 1b and a standard error of prediction (SEP) of 0.51 mM 1b in the range of 0-25 mM. This error of prediction is approximately 0.8% of the initial concentration of 1b and is well within our specifications of <2% initial concentration.