Comparing Diagnostic Accuracy of Radiologists versus GPT-4V and Gemini Pro Vision Using Image Inputs from Diagnosis Please Cases.

Background The diagnostic abilities of multimodal large language models (LLMs) using direct image inputs and the impact of the temperature parameter of LLMs remain unexplored. Purpose To investigate the ability of GPT-4V and Gemini Pro Vision in generating differential diagnoses at different temperatures compared with radiologists using Radiology Diagnosis Please cases. Materials and Methods This retrospective study included Diagnosis Please cases published from January 2008 to October 2023. Input images included original images and captures of the textual patient history and figure legends (without imaging findings) from PDF files of each case. The LLMs were tasked with providing three differential diagnoses, repeated five times at temperatures 0, 0.5, and 1. Eight subspecialty-trained radiologists solved cases. An experienced radiologist compared generated and final diagnoses, considering the result correct if the generated diagnoses included the final diagnosis after five repetitions. Accuracy was assessed across models, temperatures, and radiology subspecialties, with statistical significance set at P < .007 after Bonferroni correction for multiple comparisons across the LLMs at the three temperatures and with radiologists. Results A total of 190 cases were included in neuroradiology (n = 53), multisystem (n = 27), gastrointestinal (n = 25), genitourinary (n = 23), musculoskeletal (n = 17), chest (n = 16), cardiovascular (n = 12), pediatric (n = 12), and breast (n = 5) subspecialties. Overall accuracy improved with increasing temperature settings (0, 0.5, 1) for both GPT-4V (41% [78 of 190 cases], 45% [86 of 190 cases], 49% [93 of 190 cases], respectively) and Gemini Pro Vision (29% [55 of 190 cases], 36% [69 of 190 cases], 39% [74 of 190 cases], respectively), although there was no evidence of a statistically significant difference after Bonferroni adjustment (GPT-4V, P = .12; Gemini Pro Vision, P = .04). The overall accuracy of radiologists (61% [115 of 190 cases]) was higher than that of Gemini Pro Vision at temperature 1 (T1) (P < .001), while no statistically significant difference was observed between radiologists and GPT-4V at T1 after Bonferroni adjustment (P = .02). Radiologists (range, 45%-88%) outperformed the LLMs at T1 (range, 24%-75%) in most subspecialties. Conclusion Using direct radiologic image inputs, GPT-4V and Gemini Pro Vision showed improved diagnostic accuracy with increasing temperature settings. Although GPT-4V slightly underperformed compared with radiologists, it nonetheless demonstrated promising potential as a supportive tool in diagnostic decision-making. © RSNA, 2024 See also the editorial by Nishino and Ballard in this issue.

Kinetic Modulation of Amyloid-ß (1-42) Aggregation and Toxicity by Structure-Based Rational Design.

Several point mutations can modulate protein structure and dynamics, leading to different natures. Especially in the case of amyloidogenic proteins closely related to neurodegenerative diseases, structural changes originating from point mutations can affect fibrillation kinetics. Herein, we rationally designed mutant candidates to inhibit the fibrillation process of amyloid-ß with its point mutants through multistep in silico analyses. Our results showed that the designed mutants induced kinetic self-assembly suppression and reduced the toxicity of the aggregate. A multidisciplinary biophysical approach with small-angle X-ray scattering, ion mobility-mass spectrometry, mass spectrometry, and additional in silico experiments was performed to reveal the structural basis associated with the inhibition of fibril formation. The structure-based design of the mutants with suppressed self-assembly performed in this study could provide a different perspective for modulating amyloid aggregation based on the structural understanding of the intrinsically disordered proteins.

Diagnostic performance of the medial temporal lobe atrophy scale in patients with Alzheimer's disease: a systematic review and meta-analysis.

OBJECTIVE: To evaluate the diagnostic performance and reliability of the medial temporal lobe atrophy (MTA) scale in patients with Alzheimer's disease. METHODS: A systematic literature search of MEDLINE and EMBASE databases was performed to select studies that evaluated the diagnostic performance or reliability of MTA scale, published up to January 21, 2021. Pooled estimates of sensitivity and specificity were calculated using a bivariate random-effects model. Pooled correlation coefficients for intra- and interobserver agreements were calculated using the random-effects model based on Fisher's Z transformation of correlations. Meta-regression was performed to explain the study heterogeneity. Subgroup analysis was performed to compare the diagnostic performance of the MTA scale and hippocampal volumetry. RESULTS: Twenty-one original articles were included. The pooled sensitivity and specificity of the MTA scale in differentiating Alzheimer's disease from healthy control were 74% (95% CI, 68-79%) and 88% (95% CI, 83-91%), respectively. The area under the curve of the MTA scale was 0.88 (95% CI, 0.84-0.90). Meta-regression demonstrated that the difference in the method of rating the MTA scale was significantly associated with study heterogeneity (p = 0.04). No significant difference was observed in five studies regarding the diagnostic performance between MTA scale and hippocampal volumetry (p = 0.40). The pooled correlation coefficients for intra- and interobserver agreements were 0.85 (95% CI, 0.69-0.93) and 0.83 (95% CI, 0.66-0.92), respectively. CONCLUSIONS: Our meta-analysis demonstrated a good diagnostic performance and reliability of the MTA scale in Alzheimer's disease. KEY POINTS: • The pooled sensitivity and specificity of the MTA scale in differentiating Alzheimer's disease from healthy control were 74% and 88%, respectively. • There was no significant difference in the diagnostic performance between MTA scale and hippocampal volumetry. • The reliability of MTA scale was excellent based on the pooled correlation coefficient for intra- and interobserver agreements.

Modification of the Elevation Plane and Defatting Technique to Create a Thin Thoracodorsal Artery Perforator Flap.

BACKGROUND: Perforator flaps have been used extensively in the field of reconstruction, and the thoracodorsal artery perforator (TDAP) flap is very popular. However, the perforator flap can be relatively bulky in some cases, depending on the defect's location. Thus, several methods have been developed to address this bulkiness, including modification of the flap elevation, application of an ultrathin flap using microdissection, and the defatting technique. However, these methods have various disadvantages, so we developed an adjustable thin TDAP flap using modification of the flap elevation and defatting technique. METHODS: Between January 2012 and February 2015, 13 patients underwent reconstruction of defects of their upper and lower extremities using TDAP flaps. We measured all the flap dimensions, except for thickness, because it was adjusted for the target defect. RESULTS: The mean flap size was 94 cm(2) (range: 48-210 cm(2)), and all flaps were ≤10 cm wide to facilitate primary donor-site closure. Two subjects with a history of diabetes exhibited partial flap loss, so we performed secondary skin graft surgery. CONCLUSIONS: The TDAP flap elevation was modified at the superficial fascia plane, and the defatting technique was used to adjust the flap volume. This technique provided more natural contours and minimized the need for secondary debulking.

Use of the chimeric anterolateral thigh free flap in lower extremity reconstruction.

BACKGROUND: Reconstruction of three-dimensional lower extremity defects is challenging because the dead space should be filled and the surface defect should be covered to prevent complications. We present our experience using the vastus lateralis muscle-chimeric anterolateral thigh (ALT) free flap for reconstructing three-dimensional lower extremity defects. METHODS: This report describes 12 cases of three-dimensional lower extremity defects that were treated via reconstruction using a chimeric ALT free flap between October 2010 and January 2015. The defects involved the foot (10 patients), distal lower leg (1 patient), and proximal lower leg (1 patient). The sizes of the surface defects ranged from 7.5 × 3 cm(2) to 16 × 7 cm(2), and the sizes of the estimated dead spaces ranged from 2 × 3 cm(2) to 8 × 5 cm(2). The skin and muscle segment sizes were also evaluated. RESULTS: The sizes of the skin flaps ranged from 8 × 4 cm(2) to 17.5 × 8 cm(2), and the sizes of the muscle segments ranged from 2 × 3 cm(2) to 9 × 5 cm(2). Eleven cases exhibited full flap survival and one case exhibited partial necrosis. The follow-up periods ranged from 2 months to 38 months. We did not observe any ranges of motion limitations in the hip and knee joints of the operated leg, or any secondary complications (e.g., abscess or prolonged drainage). CONCLUSIONS: The vastus lateralis muscle-chimeric ALT free flap is a useful option for reconstructing three-dimensional lower extremity defects.

Down-regulation of microglial activity attenuates axotomized nigral dopaminergic neuronal cell loss.

BACKGROUND: There is growing evidence that inflammatory processes of activated microglia could play an important role in the progression of nerve cell damage in neurodegenerative disorders such as Parkinson's disease and Alzheimer's disease which harbor features of chronic microglial activation, though the precise mechanism is unknown. In this study, we presented in vivo and ex vivo experimental evidences indicating that activated microglia could exacerbate the survival of axotomized dopaminergic neurons and that appropriate inactivation of microglia could be neuroprotective. RESULTS: The transection of medial forebrain bundle (MFB) of a rat induced loss of dopaminergic neurons in a time-dependent manner and accompanied with microglial activation. Along with microglial activation, production of reactive oxygen species (ROS) was upregulated and TH/OX6/hydroethidine triple-immunofluorescence showed that the microglia mainly produced ROS. When the activated microglial cells that were isolated from the substantia nigra of the MFB axotomized animal, were transplanted into the substantia nigra of which MFB had been transected at 7 days ago, the survival rate of axotomized dopaminergic neurons was significantly reduced as compared with sham control. Meanwhile, when the microglial activation was attenuated by administration of tuftsin fragment 1-3 (microglia inhibitory factor) into the lateral ventricle using mini-osmotic pump, the survival rate of axotomized dopaminergic neurons was increased. CONCLUSION: The present study suggests that activated microglia could actively produce and secrete unfavorable toxic substances, such as ROS, which could accelerate dopaminergic neuronal cell loss. So, well-controlled blockade of microglial activation might be neuroprotective in some neuropathological conditions.

Anesthetic modulations dissociate neuroelectric characteristics between sensory-evoked and spontaneous activities across bilateral rat somatosensory cortical laminae.

Spontaneous neural activity has been widely adopted to construct functional connectivity (FC) amongst distant brain regions. Although informative, the functional role and signaling mechanism of the resting state FC are not intuitive as those in stimulus/task-evoked activity. In order to bridge the gap, we investigated anesthetic modulation of both resting-state and sensory-evoked activities. We used two well-studied GABAergic anesthetics of varying dose (isoflurane: 0.5-2.0% and α-chloralose: 30 and 60 mg/kg∙h) and recorded changes in electrophysiology using a pair of laminar electrode arrays that encompass the entire depth of the bilateral somatosensory cortices (S1fl) in rats. Specifically, the study focused to describe how varying anesthesia conditions affect the resting state activities and resultant FC between bilateral hemispheres in comparison to those obtained by evoked responses. As results, isoflurane decreased the amplitude of evoked responses in a dose-dependent manner mostly due to the habituation of repetitive responses. However, α-chloralose rather intensified the amplitude without exhibiting habituation. No such diverging trend was observed for the spontaneous activity, in which both anesthetics increased the signal power. For α-chloralose, overall FC was similar to that obtained with the lowest dose of isoflurane at 0.5% while higher doses of isoflurane displayed increased FC. Interestingly, only α-chloralose elicited relatively much greater increases in the ipsi-stimulus evoked response (i.e., in S1fl ipsilateral to the stimulated forelimb) than those associated with the contra-stimulus response, suggesting enhanced neuronal excitability. Taken together, the findings demonstrate modulation of the FC profiles by anesthesia is highly non-linear, possibly with a distinct underlying mechanism that affects either resting state or evoked activities differently. Further, the current study warrants thorough investigation of the basal neuronal states prior to the interpretation of resting state FC and evoked activities for accurate understanding of neural signal processing and circuitry.

Direct observation of protein structural transitions through entire amyloid aggregation processes in water using 2D-IR spectroscopy.

Amyloid proteins that undergo self-assembly to form insoluble fibrillar aggregates have attracted much attention due to their role in biological and pathological significance in amyloidosis. This study aims to understand the amyloid aggregation dynamics of insulin (INS) in H2O using two-dimensional infrared (2D-IR) spectroscopy. Conventional IR studies have been performed in D2O to avoid spectral congestion despite distinct H-D isotope effects. We observed a slowdown of the INS fibrillation process in D2O compared to that in H2O. The 2D-IR results reveal that different quaternary structures of INS at the onset of the nucleation phase caused the distinct fibrillation pathways of INS in H2O and D2O. A few different biophysical analysis, including solution-phase small-angle X-ray scattering combined with molecular dynamics simulations and other spectroscopic techniques, support our 2D-IR investigation results, providing insight into mechanistic details of distinct structural transition dynamics of INS in water. We found the delayed structural transition in D2O is due to the kinetic isotope effect at an early stage of fibrillation of INS in D2O, i.e., enhanced dimer formation of INS in D2O. Our 2D-IR and biophysical analysis provide insight into mechanistic details of structural transition dynamics of INS in water. This study demonstrates an innovative 2D-IR approach for studying protein dynamics in H2O, which will open the way for observing protein dynamics under biological conditions without IR spectroscopic interference by water vibrations.

TEMPO-Assisted Free-Radical-Initiated Peptide Sequencing Mass Spectrometry for Ubiquitin Ions: An Insight on the Gas-Phase Conformations.

TEMPO ((2,2,6,6-tetramethylpiperidine-1-yl)oxyl)-assisted free-radical-initiated peptide sequencing mass spectrometry (FRIPS MS) is applied to the top-down tandem mass spectrometry of guanidinated ubiquitin (UB(Gu)) ions, i.e., p-TEMPO-Bn-Sc-guanidinated ubiquitin (UBT(Gu)), to shed a light on gas-phase ubiquitin conformations. Thermal activation of UBT(Gu) ions produced protein backbone fragments of radical character, i.e., a-/x- and c-/z-type fragments. It is in contrast to the collision-induced dissociation (CID) results for UB(Gu), which dominantly showed the specific charge-remote CID fragments of b-/y-type at the C-terminal side of glutamic acid (E) and aspartic acid (D). The transfer of a radical "through space" was mainly observed for the +5 and +6 UBT(Gu) ions. This provides the information about folding/unfolding and structural proximity between the positions of the incipient benzyl radical site and fragmented sites. The analysis of FRIPS MS results for the +5 charge state ubiquitin ions shows that the +5 charge state ubiquitin ions bear a conformational resemblance to the native ubiquitin (X-ray crystallography structure), particularly in the central sequence region, whereas some deviations were observed in the unstable second structure region (ß2) close to the N-terminus. The ion mobility spectrometry results also corroborate the FRIPS MS results in terms of their conformations (or structures). The experimental results obtained in this study clearly demonstrate a potential of the TEMPO-assisted FRIPS MS as one of the methods for the elucidation of the overall gas-phase protein structures.

Development and validation of a deep-learning-based pediatric early warning system: A single-center study.

BACKGROUND: Early detection and prompt intervention for clinically deteriorating events are needed to improve clinical outcomes. There have been several attempts at this, including the introduction of rapid response teams (RRTs) with early warning scores. We developed a deep-learning-based pediatric early warning system (pDEWS) and validated its performance. METHODS: This single-center retrospective observational cohort study reviewed, 50,019 pediatric patients admitted to the general ward in a tertiary-care academic children's hospital from January 2012 to December 2018. They were split by admission date into a derivation and a validation cohort. We developed a pDEWS for the early prediction of cardiopulmonary arrest and unexpected ward-to-pediatric intensive care unit (PICU) transfer. Then, we validated this system by comparing modified pediatric early warning score (PEWS), random forest (RF); an ensemble model of multiple decision trees and logistic regression (LR); a statistical model that uses a logistic function. RESULTS: For predicting cardiopulmonary arrest, the pDEWS (area under the receiver operating characteristic curve (AUROC), 0.923) outperformed modified PEWS (AUROC, 0.769) and reduced the mean alarm count per day (MACPD) and number needed to examine (NNE) by 82.0% (from 46.7 to 8.4 MACPD) and 89.5% (from 0.303 to 0.807), respectively. Furthermore, for predicting unexpected ward-to-PICU transfer pDEWS also showed superior performance compared to existing methods. CONCLUSION: Our study showed that pDEWS was superior to the modified PEWS and prediction models using RF and LR. This study demonstrates that the integration of the pDEWS into RRTs could increase operational efficiency and improve clinical outcomes.

Effect of packing density of lipid vesicles on the Aß42 fibril polymorphism.

The aggregation of amyloid-ß 1-42 (Aß42) on lipid membranes is closely related to the pathology of Alzheimer's disease (AD). Herein, we demonstrated the effect of the packing density of lipid vesicles on the Aß42 fibrillation kinetics and fibril morphology. We used three distinct phosphatidylcholine (PC) lipids, containing different numbers of cis-double bonds in acyl chains, and therefore, a different packing density in the lipid vesicles. Our results showed that the fibrillation of Aß42 was greatly enhanced and the formed fibrils became shorter as the number of double bonds in lipids increased. Due to the low-density characteristics of dioleoyl phosphatidylcholine (DOPC), Aß42 monomers were able to interact with the hydrophobic acyl chain of lipids exposed to the aqueous phase, thereby inducing rapid fibrillation and short fibril morphologies. Furthermore, the effects of the anionic lipids dioleoyl phosphatidylserine (DOPS) and dioleoyl phosphatidylglycerol (DOPG), and mixed vesicles of DOPC/DOPS and DOPC/DOPG on Aß42 fibrillations were investigated. The tight binding of Aß42 to the lipid head groups via electrostatic interactions was able to suppress the modulation of Aß42 fibrillations compared to accelerated fibrillations on loosely packed membranes. Our proposed mechanism regarding the influence of lipid packing density on Aß42 fibrillations provides an advanced understanding of lipid-associated amyloid fibrillations.

The neuroprotective effects of pregabalin after cerebral ischemia by occlusion of the middle cerebral artery in rats.

Activation of presynaptic voltage-gated calcium channels and glutamate release serves a central role in neuronal necrosis after cerebral ischemia. Pregabalin binds to the α2-δ subunit of voltage-gated calcium channels and results in reduced glutamate release. The aim of the current study was to evaluate the effect of pregabalin on cerebral outcome following cerebral ischemia using an established rat model. Male Sprague-Dawley rats were randomized to receive oral administration of 5 mg/kg pregabalin for 1 day (PD1 group) or 5 days (PD5 group), or an equal amount of normal saline for 1 day (SD1 group) or 5 days (SD5 group) after 1 day of middle cerebral artery occlusion (MCAO) and reperfusion. Behavioral tests were assessed at postoperative days 1 and 7. Cerebral infarct volume was measured using a brain MRI scan on days 1 and 7 following surgery. Using immunohistochemistry to detect brain-derived neurotrophic factor (BDNF), histologic examinations of perilesional cortex and ipsilateral hippocampus were performed at postoperative day 7. BDNF-positive immunostaining was more abundant in the perilesional cortex of mice of the PD1 group compared with mice of the SD1 group (P=0.001). In the ipsilateral hippocampus, greater BDNF-positive staining was present in the PD5 group compared with the SD5 group (P=0.04). No statistically significant differences were indicated for behavioral tests or cerebral infarct volume between the PD1 and SD1 groups or the PD5 and SD5 groups. In conclusion, treatment with pregabalin beneficially impacts BDNF expression and histologic cerebral outcome in rats after cerebral ischemia.

Early treadmill exercise increases macrophage migration inhibitory factor expression after cerebral ischemia/reperfusion.

The neuroprotective function of macrophage migration inhibitory factor (MIF) in ischemic stroke was rarely evaluated. This study aimed to investigate the effects of early treadmill exercise on recovery from ischemic stroke and to determine whether these effects are associated with the expression levels of MIF and brain-derived neurotrophic factor (BDNF) in the ischemic area. A total of 40 male Sprague-Dawley rats were randomly assigned to the ischemia and exercise group [middle cerebral artery occlusion (MCAO)-Ex, n = 10), ischemia and sedentary group (MCAO-St, n = 10), sham-surgery and exercise group (Sham-Ex, n = 10), or sham-surgery and sedentary group (Sham-St, n = 10). The MCAO-Ex and MCAO-St groups were subjected to MCAO for 60 minutes, whereas the Sham-Ex and Sham-St groups were subjected to an identical operation without MCAO. Rats in the MCAO-Ex and Sham-Ex groups then ran on a treadmill for 30 minutes once a day for 5 consecutive days. After reperfusion, the hanging time tested by the wire hang test was longer and the relative fractional anisotropy determined by MRI was higher in the peri-infarct region of the MCAO-Ex group compared with the MCAO-St group. The expression levels of MIF and BDNF in the peri-infarct region were upregulated in the MCAO-Ex group. Increased MIF and BDNF levels were positively correlated with relative fractional anisotropy changes in the peri-infarct region. There was no significant difference in the levels of MIF and BDNF in the peri-infarct region between the Sham-Ex and Sham-St groups. Our study demonstrated that early exercise (initiated 48 hours after the MCAO) could improve motor and neuronal recovery after ischemic stroke. Furthermore, the increased levels of MIF and BDNF in the peri-infarct region (penumbra) may be one of the mechanisms of enhanced neurological function recovery. All experiments were approved by the Institutional Animal Care and Use Committee in Asan Medical Center in South Korea (2016-12-126).

Effects of anodal transcranial direct current stimulation (tDCS) on behavioral and spatial memory during the early stage of traumatic brain injury in the rats.

Transcranial direct current stimulation (tDCS) is a noninvasive technique to modulate the neural membrane potential. Its effects in the early stage of traumatic brain injury (TBI) have rarely been investigated. This study assessed the effects of anodal tDCS on behavioral and spatial memory in a rat model of traumatic brain injury. Thirty six rats underwent lateral fluid percussion and were then randomly assigned to one of three groups: control (n=12), five-day tDCS over peri-lesional cortex at one (1W, n=12), or two (2W, n=12) weeks post-injury. The Barnes maze (BM) and Rotarod (RR) tests were evaluated in a blind manner on day 1, week 3 and week 5 post-injury. After three weeks, both the 1W and 2W groups showed significant improvements in the BM ratio (P<0.05), whereas only group 2W obtained a significant improvement in the RR ratio compared with the control group (P<0.05). However, there were no significant differences between any of the groups at five weeks after TBI. Immunohistochemistry revealed that only group 2W had a significantly higher brain-derived neurotrophic factor (BDNF) expression in the peri-lesional cortex, which was significantly correlated with the improvement of the Rotarod test at 3-week post-injury. However, BDNF expression in the ipsi-lesional hippocampus was not significantly different among the three groups. Group 1W tended to have increased choline/creatine ratios, as measured by magnetic resonance spectroscopy in the peri-lesional cortex, than the control group (P=0.051). Neither regimen aggravated the lesion volume or brain edema measured by MRI. These beneficial effects were not observed with either regimen at five weeks post-injury. In conclusions, anodal tDCS ameliorated behavioral and spatial memory function in the early phase after TBI when it is delivered two weeks post-injury. Earlier stimulation (one week post-injury) improves spatial memory only. However, the beneficial effects did not persist after cessation of the anodal stimulation.

Role of activating transcription factor 3 in ischemic penumbra region following transient middle cerebral artery occlusion and reperfusion injury.

The activating transcription factor 3 (ATF3) is expressed by various types of cellular insults. It has been suggested to serve diverse functions in both cellular survival and death signal cascades, but the exact role of ATF3 in brain ischemia is little known so far. Thus, the authors examined the expression pattern of ATF3 following middle cerebral artery occlusion (MCAO) and reperfusion injury. At 1-2 days after MCAO and reperfusion injury, numerous number of ATF3-immunoreacitive (-ir) nuclei was observed in the ipsilateral peri-infarct cortex, but declined rapidly at 3 days. Almost all ATF3-ir nuclei were co-localized with NeuN-ir neurons. Neither GFAP- nor OX42-ir neuroglia were co-localized with ATF3. Double labeling of Fluoro-Jade B with ATF3 showed that ATF3-ir nuclei mismatched with Fluoro-Jade B-ir neurons. To further examine the role of ATF3 in ischemic peri-infarct regions, double immunofluorescent labeling of ATF3/caspase 3, ATF3/Bcl-xl, and ATF3/HSP27 was conducted. Semiquantitive estimation showed that about 15% of ATF3-ir neurons also expressed caspase 3. However, about only 0.4% and 2.6% of ATF3-ir neurons were double-stained with Bcl-xl and Hsp27, respectively. Consequently, it would be suggested that ATF3 seem to play an important role in caspase-dependent neuronal apoptotic signal transduction pathways caused by focal cerebral ischemia and reperfusion injury.