Category: Nevin Manimala

Artif Intell Med. 2023 Aug;142:102588. doi: 10.1016/j.artmed.2023.102588. Epub 2023 May 20.

ABSTRACT

BACKGROUND: Amyotrophic Lateral Sclerosis (ALS) is a fatal neurodegenerative disorder characterised by the progressive loss of motor neurons in the brain and spinal cord. The fact that ALS’s disease course is highly heterogeneous, and its determinants not fully known, combined with ALS’s relatively low prevalence, renders the successful application of artificial intelligence (AI) techniques particularly arduous.

OBJECTIVE: This systematic review aims at identifying areas of agreement and unanswered questions regarding two notable applications of AI in ALS, namely the automatic, data-driven stratification of patients according to their phenotype, and the prediction of ALS progression. Differently from previous works, this review is focused on the methodological landscape of AI in ALS.

METHODS: We conducted a systematic search of the Scopus and PubMed databases, looking for studies on data-driven stratification methods based on unsupervised techniques resulting in (A) automatic group discovery or (B) a transformation of the feature space allowing patient subgroups to be identified; and for studies on internally or externally validated methods for the prediction of ALS progression. We described the selected studies according to the following characteristics, when applicable: variables used, methodology, splitting criteria and number of groups, prediction outcomes, validation schemes, and metrics.

RESULTS: Of the starting 1604 unique reports (2837 combined hits between Scopus and PubMed), 239 were selected for thorough screening, leading to the inclusion of 15 studies on patient stratification, 28 on prediction of ALS progression, and 6 on both stratification and prediction. In terms of variables used, most stratification and prediction studies included demographics and features derived from the ALSFRS or ALSFRS-R scores, which were also the main prediction targets. The most represented stratification methods were K-means, and hierarchical and expectation-maximisation clustering; while random forests, logistic regression, the Cox proportional hazard model, and various flavours of deep learning were the most widely used prediction methods. Predictive model validation was, albeit unexpectedly, quite rarely performed in absolute terms (leading to the exclusion of 78 eligible studies), with the overwhelming majority of included studies resorting to internal validation only.

CONCLUSION: This systematic review highlighted a general agreement in terms of input variable selection for both stratification and prediction of ALS progression, and in terms of prediction targets. A striking lack of validated models emerged, as well as a general difficulty in reproducing many published studies, mainly due to the absence of the corresponding parameter lists. While deep learning seems promising for prediction applications, its superiority with respect to traditional methods has not been established; there is, instead, ample room for its application in the subfield of patient stratification. Finally, an open question remains on the role of new environmental and behavioural variables collected via novel, real-time sensors.

PMID:37316101 | DOI:10.1016/j.artmed.2023.102588

Artif Intell Med. 2023 Aug;142:102587. doi: 10.1016/j.artmed.2023.102587. Epub 2023 May 22.

ABSTRACT

OBJECTIVE: The proper handling of missing values is critical to delivering reliable estimates and decisions, especially in high-stakes fields such as clinical research. In response to the increasing diversity and complexity of data, many researchers have developed deep learning (DL)-based imputation techniques. We conducted a systematic review to evaluate the use of these techniques, with a particular focus on the types of data, intending to assist healthcare researchers from various disciplines in dealing with missing data.

MATERIALS AND METHODS: We searched five databases (MEDLINE, Web of Science, Embase, CINAHL, and Scopus) for articles published prior to February 8, 2023 that described the use of DL-based models for imputation. We examined selected articles from four perspectives: data types, model backbones (i.e., main architectures), imputation strategies, and comparisons with non-DL-based methods. Based on data types, we created an evidence map to illustrate the adoption of DL models.

RESULTS: Out of 1822 articles, a total of 111 were included, of which tabular static data (29%, 32/111) and temporal data (40%, 44/111) were the most frequently investigated. Our findings revealed a discernible pattern in the choice of model backbones and data types, for example, the dominance of autoencoder and recurrent neural networks for tabular temporal data. The discrepancy in imputation strategy usage among data types was also observed. The “integrated” imputation strategy, which solves the imputation task simultaneously with downstream tasks, was most popular for tabular temporal data (52%, 23/44) and multi-modal data (56%, 5/9). Moreover, DL-based imputation methods yielded a higher level of imputation accuracy than non-DL methods in most studies.

CONCLUSION: The DL-based imputation models are a family of techniques, with diverse network structures. Their designation in healthcare is usually tailored to data types with different characteristics. Although DL-based imputation models may not be superior to conventional approaches across all datasets, it is highly possible for them to achieve satisfactory results for a particular data type or dataset. There are, however, still issues with regard to portability, interpretability, and fairness associated with current DL-based imputation models.

PMID:37316097 | DOI:10.1016/j.artmed.2023.102587

Artif Intell Med. 2023 Aug;142:102555. doi: 10.1016/j.artmed.2023.102555. Epub 2023 Apr 28.

ABSTRACT

Digital mammography is currently the most common imaging tool for breast cancer screening. Although the benefits of using digital mammography for cancer screening outweigh the risks associated with the x-ray exposure, the radiation dose must be kept as low as possible while maintaining the diagnostic utility of the generated images, thus minimizing patient risks. Many studies investigated the feasibility of dose reduction by restoring low-dose images using deep neural networks. In these cases, choosing the appropriate training database and loss function is crucial and impacts the quality of the results. In this work, we used a standard residual network (ResNet) to restore low-dose digital mammography images and evaluated the performance of several loss functions. For training purposes, we extracted 256,000 image patches from a dataset of 400 images of retrospective clinical mammography exams, where dose reduction factors of 75% and 50% were simulated to generate low and standard-dose pairs. We validated the network in a real scenario by using a physical anthropomorphic breast phantom to acquire real low-dose and standard full-dose images in a commercially available mammography system, which were then processed through our trained model. We benchmarked our results against an analytical restoration model for low-dose digital mammography. Objective assessment was performed through the signal-to-noise ratio (SNR) and the mean normalized squared error (MNSE), decomposed into residual noise and bias. Statistical tests revealed that the use of the perceptual loss (PL4) resulted in statistically significant differences when compared to all other loss functions. Additionally, images restored using the PL4 achieved the closest residual noise to the standard dose. On the other hand, perceptual loss PL3, structural similarity index (SSIM) and one of the adversarial losses achieved the lowest bias for both dose reduction factors. The source code of our deep neural network is available at https://github.com/WANG-AXIS/LdDMDenoising.

PMID:37316093 | DOI:10.1016/j.artmed.2023.102555

Food Res Int. 2023 Aug;170:113027. doi: 10.1016/j.foodres.2023.113027. Epub 2023 May 24.

ABSTRACT

Mercury (Hg) was reported to accumulate in rice grains, and, together with the selenium (Se) was found in rice, the co-exposure of Hg-Se via rice consumption may present significant health effects to human. This research collected rice samples containing high Hg:high Se and high Se:low Hg concentrations from high Hg and high Se background areas. The physiologically based extraction test (PBET) in vitro digestion model was utilized to obtain bioaccessibility data from samples. The results showed relatively low bioaccessible for Hg (<60%) and Se (<25%) in both rice sample groups, and no statistically significant antagonism was identified. However, the correlations of Hg and Se bioaccessibility showed an inverse pattern for the two sample groups. A negative correlation was detected in the high Se background rice group and a positive correlation in the high Hg background group, suggesting various micro forms of Hg and Se in rice from different planting locations. In addition, when the benefit-risk value (BRV) was calculated, some “fake” positive results showed while Hg and Se concentrations were directly used, which indicated that bioaccessibility should not be neglected in benefit-risk assessment.

PMID:37316027 | DOI:10.1016/j.foodres.2023.113027

Food Res Int. 2023 Aug;170:113015. doi: 10.1016/j.foodres.2023.113015. Epub 2023 May 21.

ABSTRACT

Accurately and high-thoroughly screening illegal additives in health-care foods continues to be a challenging task in routine analysis for the ultrahigh performance liquid chromatography-high resolution mass spectrometry based techniques. In this work, we proposed a new strategy to identify additives in complex food matrices, which consists of both experimental design and advanced chemometric data analysis. At first, reliable features in the analyzed samples were screened based on a simple but efficient sample weighting design, and those related to illegal additives were screened with robust statistical analysis. After the MS¹ in-source fragment ion identification, both MS¹ and MS/MS spectra were constructed for each underlying compound, based on which illegal additives can be precisely identified. The performance of the developed strategy was demonstrated by using mixture and synthetic sample datasets, indicating an improvement of data analysis efficiency up to 70.3 %. Finally, the developed strategy was applied for the screening of unknown additives in 21 batches of commercially available health-care foods. Results indicated that at least 80 % of false-positive results can be reduced and 4 additives were screened and confirmed.

PMID:37316023 | DOI:10.1016/j.foodres.2023.113015

Int J Spine Surg. 2023 Jun 14:8454. doi: 10.14444/8454. Online ahead of print.

ABSTRACT

BACKGROUND: Durotomy during endoscopic spine surgery can cause a patient’s neurological or cardiovascular status to deteriorate unexpectedly intra- or postoperatively. There is currently limited literature regarding appropriate fluid management strategies, irrigation-related risk factors, and clinical consequences of incidental durotomy during spinal endoscopy, and no validated irrigation protocol exists for endoscopic spine surgery. Thus, the present article sought to (1) describe 3 cases of durotomy, (2) investigate standard epidural pressure measurements, and (3) survey endoscopic spine surgeons on the incidence of adverse effects believed to result from durotomy.

MATERIALS AND METHODS: The authors first reviewed clinical outcomes and analyzed complications in 3 patients with intraoperatively recognized incidental durotomy. Second, the authors conducted a small case series with intraoperative epidural pressure measurements during gravity-assisted irrigated video endoscopy of the lumbar spine. Measurements were conducted on 12 patients with a transducer assembly that was introduced through the endoscopic working channel of the RIWOSpine Panoview Plus and Vertebris endoscope to the decompression site in the spine. Third, the authors conducted a retrospective, multiple-choice survey of endoscopic spine surgeons to better understand the frequency and seriousness of problems they attributed to irrigation fluid escaping from the surgical decompression site into the spinal canal and neural axis. Descriptive and correlative statistical analyses were performed on the surgeons’ responses.

RESULTS: In the first part of this study, durotomy-related complications during irrigated spinal endoscopy were observed in 3 patients. Postoperative head computed tomographic (CT) images revealed massive blood in the intracranial subarachnoid space, the basal cisterns, the III and IV ventricle, and the lateral ventricles characteristic of an arterial fisher grade IV subarachnoid hemorrhage, and hydrocephalus without evidence of aneurysms or angiomas. Two additional patients developed intraoperative seizures, cardiac arrhythmia, and hypotension. The head CT image in 1 of these 2 patients had intracranial air entrapment.In the second part, epidural pressure measurements in 12 patients who underwent uneventful routine lumbar interlaminar decompression for L4-L5 and L5-S1 disc herniation showed an average epidural pressure of 24.5 mm Hg.In the third part, the online survey was accessed by 766 spine surgeons worldwide and had a response rate of 43.6%. Irrigation-related problems were reported by 38% of responding surgeons. Only 11.8% used irrigation pumps, with 90% running the pump above 40 mm Hg. Headaches (4.5%) and neck pain (4.9%) were observed by nearly a 10th (9.4%) of surgeons. Seizures in combination with headaches, neck and abdominal pain, soft tissue edema, and nerve root injury were reported by another 5 surgeons. One surgeon reported a delirious patient. Another 14 surgeons thought that they had patients with neurological deficits ranging from nerve root injury to cauda equina syndrome related to irrigation fluid. Autonomic dysreflexia associated with hypertension was attributed by 19 of the 244 responding surgeons to the noxious stimulus of escaped irrigation fluid that migrated from the decompression site in the spinal canal. Two of these 19 surgeons reported 1 case associated with a recognized incidental durotomy and another with postoperative paralysis.

CONCLUSIONS: Patients should be educated preoperatively about the risk of irrigated spinal endoscopy. Although rare, intracranial blood, hydrocephalus, headaches, neck pain, seizures, and more severe complications, including life-threatening autonomic dysreflexia with hypertension, may arise if irrigation fluid enters the spinal canal or the dural sac and migrates from the endoscopic site along the neural axis rostrally. Experienced endoscopic spine surgeons suspect a correlation between durotomy and irrigation-related extra- and intradural pressure equalization that could be problematic if associated with high volumes of irrigation fluid LEVEL OF EVIDENCE: 3.

PMID:37315993 | DOI:10.14444/8454

BMJ Evid Based Med. 2023 Jun 14:bmjebm-2023-112292. doi: 10.1136/bmjebm-2023-112292. Online ahead of print.

NO ABSTRACT

PMID:37315987 | DOI:10.1136/bmjebm-2023-112292

BMJ. 2023 Jun 14;381:e073843. doi: 10.1136/bmj-2022-073843.

ABSTRACT

OBJECTIVE: To determine the effectiveness of risk stratification using the Global Registry of Acute Coronary Events (GRACE) risk score (GRS) for patients presenting to hospital with suspected non-ST elevation acute coronary syndrome.

DESIGN: Parallel group cluster randomised controlled trial.

SETTING: Patients presenting with suspected non-ST elevation acute coronary syndrome to 42 hospitals in England between 9 March 2017 and 30 December 2019.

PARTICIPANTS: Patients aged ≥18 years with a minimum follow-up of 12 months.

INTERVENTION: Hospitals were randomised (1:1) to patient management by standard care or according to the GRS and associated guidelines.

MAIN OUTCOME MEASURES: Primary outcome measures were use of guideline recommended management and time to the composite of cardiovascular death, non-fatal myocardial infarction, new onset heart failure hospital admission, and readmission for cardiovascular event. Secondary measures included the duration of hospital stay, EQ-5D-5L (five domain, five level version of the EuroQoL index), and the composite endpoint components.

RESULTS: 3050 participants (1440 GRS, 1610 standard care) were recruited in 38 UK clusters (20 GRS, 18 standard care). The mean age was 65.7 years (standard deviation 12), 69% were male, and the mean baseline GRACE scores were 119.5 (standard deviation 31.4) and 125.7 (34.4) for GRS and standard care, respectively. The uptake of guideline recommended processes was 77.3% for GRS and 75.3% for standard care (odds ratio 1.16, 95% confidence interval 0.70 to 1.92, P=0.56). The time to the first composite cardiac event was not significantly improved by the GRS (hazard ratio 0.89, 95% confidence interval 0.68 to 1.16, P=0.37). Baseline adjusted EQ-5D-5L utility at 12 months (difference -0.01, 95% confidence interval -0.06 to 0.04) and the duration of hospital admission within 12 months (mean 11.2 days, standard deviation 18 days v 11.8 days, 19 days) were similar for GRS and standard care.

CONCLUSIONS: In adults presenting to hospital with suspected non-ST elevation acute coronary syndrome, the GRS did not improve adherence to guideline recommended management or reduce cardiovascular events at 12 months.

TRIAL REGISTRATION: ISRCTN 29731761.

PMID:37315959 | DOI:10.1136/bmj-2022-073843

Stat Med. 2023 Jun 14. doi: 10.1002/sim.9755. Online ahead of print.

ABSTRACT

Existing statistical methods can estimate a policy, or a mapping from covariates to decisions, which can then instruct decision makers (eg, whether to administer hypotension treatment based on covariates blood pressure and heart rate). There is great interest in using such data-driven policies in healthcare. However, it is often important to explain to the healthcare provider, and to the patient, how a new policy differs from the current standard of care. This end is facilitated if one can pinpoint the aspects of the policy (ie, the parameters for blood pressure and heart rate) that change when moving from the standard of care to the new, suggested policy. To this end, we adapt ideas from Trust Region Policy Optimization (TRPO). In our work, however, unlike in TRPO, the difference between the suggested policy and standard of care is required to be sparse, aiding with interpretability. This yields “relative sparsity,” where, as a function of a tuning parameter, λ $$ lambda $$ , we can approximately control the number of parameters in our suggested policy that differ from their counterparts in the standard of care (eg, heart rate only). We propose a criterion for selecting λ $$ lambda $$ , perform simulations, and illustrate our method with a real, observational healthcare dataset, deriving a policy that is easy to explain in the context of the current standard of care. Our work promotes the adoption of data-driven decision aids, which have great potential to improve health outcomes.

PMID:37315949 | DOI:10.1002/sim.9755

Stat Med. 2023 Jun 14. doi: 10.1002/sim.9825. Online ahead of print.

ABSTRACT

There has been growing research interest in developing methodology to evaluate healthcare centers’ performance with respect to patient outcomes. Conventional assessments can be conducted using fixed or random effects models, as seen in provider profiling. We propose a new method, using fusion penalty to cluster healthcare centers with respect to a survival outcome. Without any prior knowledge of the grouping information, the new method provides a desirable data-driven approach for automatically clustering healthcare centers into distinct groups based on their performance. An efficient alternating direction method of multipliers algorithm is developed to implement the proposed method. The validity of our approach is demonstrated through simulation studies, and its practical application is illustrated by analyzing data from the national kidney transplant registry.

PMID:37315935 | DOI:10.1002/sim.9825