Skin detectors play a crucial role in many applications: face localization, person tracking, objectionable content screening, etc. Skin detection is a complicated process that involves not only the development of apposite classifiers but also many ancillary methods, including techniques for data preprocessing and postprocessing. In this paper, a new postprocessing method is described that learns to select whether an image needs the application of various morphological sequences or a homogeneity function. The type of postprocessing method selected is learned based on categorizing the image into one of eleven predetermined classes. The novel postprocessing method presented here is evaluated on ten datasets recommended for fair comparisons that represent many skin detection applications. The results show that the new approach enhances the performance of the base classifiers and previous works based only on learning the most appropriate morphological sequences.

Monitoring activities of daily living (ADLs) plays an important role in measuring and responding to a person\'s ability to manage their basic physical needs. Effective recognition systems for monitoring ADLs must successfully recognize naturalistic activities that also realistically occur at infrequent intervals. However, existing systems primarily focus on either recognizing more separable, controlled activity types or are trained on balanced datasets where activities occur more frequently. In our work, we investigate the challenges associated with applying machine learning to an imbalanced dataset collected from a fully in-the-wild environment. This analysis shows that the combination of preprocessing techniques to increase recall and postprocessing techniques to increase precision can result in more desirable models for tasks such as ADL monitoring. In a user-independent evaluation using in-the-wild data, these techniques resulted in a model that achieved an event-based F1-score of over 0.9 for brushing teeth, combing hair, walking, and washing hands. This work tackles fundamental challenges in machine learning that will need to be addressed in order for these systems to be deployed and reliably work in the real world.

The development of innovative production processes and the optimization of photobioreactors play an important role in generating industrial competitive production technologies for phototrophic biofilms. With emerse photobioreactors a technology was introduced that allowed efficient surface attached cultivation of terrestrial cyanobacteria. However, the productivity of emerse photobioreactors depends on the available cultivation surface. By the implementation of biocarriers to the bioreactor volume, the cultivation surface can be increased which potentially improves productivity and thus the production of valuable compounds. To investigate the surface attached cultivation on biocarriers new photobioreactors need to be developed. Additive manufacturing (AM) offers new opportunities for the design of photobioreactors but producing the needed transparent parts can be challenging using AM techniques. In this study an emerse fixed bed photobioreactor was designed for the use of biocarriers and manufactured using different AM processes. To validate the suitability of the photobioreactor for phototrophic cultivation, the optical properties of three-dimensional (3D)-printed transparent parts and postprocessing techniques to improve luminous transmittance of the components were investigated. We found that stereolithography 3D printing can produce parts with a high luminous transmittance of over 85% and that optimal postprocessing by sanding and clear coating improved the clarity and transmittance to more than 90%. Using the design freedom of AM resulted in a bioreactor with reduced part count and improved handling. In summary, we found that modern 3D-printing technologies and materials are suitable for the manufacturing of functional photobioreactor prototypes.

OBJECTIVE: The purpose of this study is to develop accurate and automated detection and segmentation methods for brain tumors, given their significant fatality rates, with aggressive malignant tumors like Glioblastoma Multiforme (GBM) having a five-year survival rate as low as 5 to 10%. This underscores the urgent need to improve diagnosis and treatment outcomes through innovative approaches in medical imaging and deep learning techniques.
METHODS: In this work, we propose a novel approach utilizing the two-headed UNetEfficientNets model for simultaneous segmentation and classification of brain tumors from Magnetic Resonance Imaging (MRI) images. The model combines the strengths of EfficientNets and a modified two-headed Unet model. We utilized a publicly available dataset consisting of 3064 brain MR images classified into three tumor classes: Meningioma, Glioma, and Pituitary. To enhance the training process, we performed 12 types of data augmentation on the training dataset. We evaluated the methodology using six deep learning models, ranging from UNetEfficientNet-B0 to UNetEfficientNet-B5, optimizing the segmentation and classification heads using binary cross entropy (BCE) loss with Dice and BCE with focal loss, respectively. Post-processing techniques such as connected component labeling (CCL) and ensemble models were applied to improve segmentation outcomes.
RESULTS: The proposed UNetEfficientNet-B4 model achieved outstanding results, with an accuracy of 99.4% after postprocessing. Additionally, it obtained high scores for DICE (94.03%), precision (98.67%), and recall (99.00%) after post-processing. The ensemble technique further improved segmentation performance, with a global DICE score of 95.70% and Jaccard index of 91.20%.
CONCLUSIONS: Our study demonstrates the high efficiency and accuracy of the proposed UNetEfficientNet-B4 model in the automatic and parallel detection and segmentation of brain tumors from MRI images. This approach holds promise for improving diagnosis and treatment planning for patients with brain tumors, potentially leading to better outcomes and prognosis.

Relative cerebral blood volume (rCBV) derived from dynamic susceptibility contrast (DSC) perfusion MR imaging (pMRI) has been shown to be a robust marker of neuroradiological tumor burden. Recent consensus recommendations in pMRI acquisition strategies have provided a pathway for pMRI inclusion in diverse patient care centers, regardless of size or experience. However, even with proper implementation and execution of the DSC-MRI protocol, issues will arise that many centers may not easily recognize or be aware of. Furthermore, missed pMRI issues are not always apparent in the resulting rCBV images, potentiating inaccurate or missed radiological diagnoses. Therefore, we gathered from our database of DSC-MRI datasets, true-to-life examples showcasing the breakdowns in acquisition, postprocessing, and interpretation, along with appropriate mitigation strategies when possible. The pMRI issues addressed include those related to image acquisition and postprocessing with a focus on contrast agent administration, timing, and rate, signal-to-noise quality, and susceptibility artifact. The goal of this work is to provide guidance to minimize and recognize pMRI issues to ensure that only quality data is interpreted.

Agitation is one of the most prevalent symptoms in people with dementia (PwD) that can place themselves and the caregiver\'s safety at risk. Developing objective agitation detection approaches is important to support health and safety of PwD living in a residential setting. In a previous study, we collected multimodal wearable sensor data from 17 participants for 600 days and developed machine learning models for detecting agitation in 1-min windows. However, there are significant limitations in the dataset, such as imbalance problem and potential imprecise labels as the occurrence of agitation is much rarer in comparison to the normal behaviours. In this paper, we first implemented different undersampling methods to eliminate the imbalance problem, and came to the conclusion that only 20% of normal behaviour data were adequate to train a competitive agitation detection model. Then, we designed a weighted undersampling method to evaluate the manual labeling mechanism given the ambiguous time interval assumption. After that, the postprocessing method of cumulative class re-decision (CCR) was proposed based on the historical sequential information and continuity characteristic of agitation, improving the decision-making performance for the potential application of agitation detection system. The results showed that a combination of undersampling and CCR improved F1-score and other metrics to varying degrees with less training time and data.
UNASSIGNED: The online version contains supplementary material available at 10.1007/s13534-023-00313-8.

Diagnosing recent small subcortical infarcts (RSSIs) via early computed tomography (CT) remains challenging. This study aimed to assess CT attenuation values (Hounsfield Units (HU)) and net water uptake (NWU) in RSSI and explore a postprocessing algorithm\'s potential to enhance thalamic RSSI detection. We examined non-contrast CT (NCCT) data from patients with confirmed thalamic RSSI on diffusion-weighted magnetic resonance imaging (DW-MRI) between January 2010 and October 2017. Co-registered DW-MRI and NCCT images enabled HU and NWU quantification in the infarct area compared to unaffected contralateral tissue. Results were categorized based on symptom onset to NCCT timing. Postprocessing using window optimization and frequency-selective non-linear blending (FSNLB) was applied, with interpretations by three blinded Neuroradiologists. The study included 34 patients (median age 70 years [IQR 63-76], 14 women). RSSI exhibited significantly reduced mean CT attenuation compared to unaffected thalamus (29.6 HU (±3.1) vs. 33.3 HU (±2.6); p < 0.01). Mean NWU in the infarct area increased from 6.4% (±7.2) at 0-6 h to 16.6% (±8.7) at 24-36 h post-symptom onset. Postprocessed NCCT using these HU values improved sensitivity for RSSI detection from 32% in unprocessed CT to 41% in FSNLB-optimized CT, with specificities ranging from 86% to 95%. In conclusion, CT attenuation values and NWU are discernible in thalamic RSSI up to 36 h post-symptom onset. Postprocessing techniques, particularly window optimization and FSNLB, moderately enhance RSSI detection.

Support removal is one of the thorny issues faced by laser powder bed fusion (LPBF). In particular, the efficient and safe removal of support structures from the thin-walled parts and obtaining high-quality surfaces still remains a challenge owing to their sensitivity to machining. An in-depth understanding of the material response behavior of LPBF thin-walled parts when removing support structures is necessary for overcoming this challenge. The work is divided into two parts: revealing the support removal mechanism and proposing a solution to improve the support machinability. First, the machinability of support structures on thin-walled parts with different thicknesses at different cutting depths was thoroughly investigated. Experimental investigation on cutting force, surface morphology, and deflection were carried out. The results show that cutting forces increase gradually at each cut owing to the tilt and collapse of support structures. The surface morphology is improved as the sample thickness increases but deteriorated as the cutting depth increases. Second, a novel solution of adding resin is proposed to improve the support machinability and good results have been achieved. The z-direction cutting forces for 0.3 and 0.4 mm thickness samples are reduced by 72.6% and 64.6%, respectively, and no deflection of the sample is observed after support removal. Moreover, finite element method simulations are established to further explain the support removal mechanism.

The challenges in reliably removing the sacrificial material from fully enclosed microfluidic channels hinder the use of three-dimensional (3D) printing to create microfluidic devices with intricate geometries. With advances in printer resolution, the etching of sacrificial materials from increasingly smaller channels is poised to be a bottleneck using the existing techniques. In this study, we introduce a microfabrication approach that utilizes centrifugation to effortlessly and efficiently remove the sacrificial materials from 3D-printed microfluidic devices with densely packed microfeatures. We characterize the process by measuring the etch rate under different centrifugal forces and developed a theoretical model to estimate process parameters for a given geometry. The effect of the device layout on the centrifugal etching process is also investigated. We demonstrate the applicability of our approach on devices fabricated using inkjet 3D printing and stereolithography. Finally, the advantages of the introduced approach over commonly used injection-based etching of sacrificial material are experimentally demonstrated in direct comparisons. A robust method to postprocess additively manufactured geometries composed of intricate microfluidic channels can help utilize both the large printing volume and high spatial resolution afforded by 3D printing in creating a variety of devices ranging from scaffolds to large-scale microfluidic assays.

OBJECTIVE: Focal cortical dysplasia (FCD) is one of the most common causes of drug-resistant epilepsy, and necessitates a multimodal evaluation to ensure optimal surgical treatment. This study aimed to determine the supportive value of the morphometric analysis program (MAP) in detecting FCD using data from a single institution in Korea.
METHODS: To develop a standard reference for the MAP, normal-looking MRIs by two scanners that are frequently used in this center were chosen. Patients with drug-resistant epilepsy and FCD after surgery were candidates for the analysis. The three-dimensional T1-weighted MRI scans of the patients were analyzed as test cases using the MAP.
RESULTS: The MRI scans of 87 patients were included in the analysis. The radiologist detected abnormal findings correlated with FCD (RAD positive [RAD(+)]) in 34 cases (39.1%), while the MAP could detect FCD in 25.3% of cases. A combination of the MAP (MAP[+] cases) with interpretations by the radiologist increased the detection to 42.5% (37 cases). The lesion detection rate was not different according to the type of reference scanners except in one case. MAP(+)/RAD(-) presented in three cases, all of which had FCD type IIa. The detection rate was slightly higher using the same kind of scanner as a reference, but not significantly (35.0% vs. 22.4% p=0.26).
CONCLUSIONS: The results of postprocessing in the MAP for detecting FCD did not depend on the type of reference scanner, and the MAP was the strongest in detecting FCD IIa. We suggested that the MAP could be widely utilized without developing institutional standards and could become an effective tool for detecting FCD lesions.