超分辨率经颅超声成像在颅内出血血肿检出和测量中的应用研究OA

Objective To explore the application value of super-resolution transcranial sonography(TCS)imaging in the detection and measurement of hematomas in intracranial hemorrhage(ICH).Methods Prospectively enroll consecutive ICH patients admitted to the intensive care unit of Department of Neurosurgery,the Second Affiliated Hospital of Air Force Medical University,from March 2025 to June 2025.Baseline and clinical data were collected,including age,sex,type of ICH(hypertensive ICH,aneurysmal ICH),diabetes mellitus,hypertension,previous cerebral infarction,coronary heart disease,admission systolic and diastolic blood pressure,admission Glasgow coma scale(GCS)score,CT to super-resolution TCS imaging time,whether surgical treatment was performed before super-resolution TCS imaging and,if so,the surgical modality(external ventricular drainage,craniotomy for intracranial hematoma evacuation,percutaneous intracranial hematoma drainage,decompressive craniectomy,craniotomy for aneurysm clipping,or endovascular embolization).All patients underwent head CT within 12 h of admission to evaluate the hemorrhage location,number of hematomas,hematoma morphology,maximum anteroposterior and mediolateral diameters of the hematoma on the axial plane,presence of perilesional edema,and whether the hematoma had ruptured into the ventricle,subarachnoid hemorrhage,multiple confluent hematomas,or diffuse hemorrhage.All patients underwent super-resolution TCS imaging within 24 h after cranial CT examination.With the three planes of the midbrain,thalamus,and lateral ventricles as reference,the probe was moved to perform continuous scanning for intracranial hematomas.Abnormal intracranial echo masses were marked as hematomas,and the location of each hematoma was determined according to the visualized anatomical landmarks.The cases with non-visualized hematomas were marked as"not detected".Hematomas that could not be reliably quantified were excluded,including those with unclear boundaries caused by diffuse hemorrhage or excessively small size(diameter≤1 cm),those with irregular morphology caused by fusion of multiple hematomas,those located in the frontal,occipital,and parietal lobes for which the full contour could not be visualized due to artifact interference,and those whose boundaries could not be accurately determined due to obvious surrounding edema.For the remaining hematomas,the slice displaying the maximum size of a hematoma was identified,and the maximum anteroposterior and mediolateral diameters of the hematoma were measured.Using head CT findings as the diagnostic standard,the results of super-resolution TCS imaging were compared with those of head CT.Preoperative super-resolution TCS imaging findings were compared with preoperative CT findings,and postoperative super-resolution TCS imaging findings were compared with postoperative CT findings.A 2 × 2 contingency table was constructed to calculate the numbers of true positives,false positives,true negatives,and false negatives for super-resolution TCS imaging,based on which the sensitivity,accuracy,and positive predictive value were calculated.Inter-observer consistency and consistency of measurements between the two imaging methods were evaluated using intraclass correlation coefficients(ICC)and Bland-Altman analysis,with ICC＞0.90 as excellent consistency,ICC＞0.75-0.90 as good consistency,ICC between 0.50 and 0.75 as moderate consistency,ICC＜0.50 as insufficient consistency;Bland-Altman analysis shows good consistency when data points was within the 95％limits of agreement(LOA).Pearson correlation analysis was used to analyze the correlation of measurements between the two imaging methods,with|r|＞0.90-1.00 indicating extremely strong correlation and 0.70-0.89 indicating strong correlation.The relative error was calculated as(|super-resolution TCS imaging measurement-CT measurement|)/CT measurement × 100％to assess the level of measurement error between the two imaging methods;a smaller relative error indicated smaller measurement deviation and higher measurement precision between the two imaging methods.Results A total of 39 patients with ICH(29 males and 10 females)were enrolled,with an age range of 25 to 83 years and a mean age of(58±13)years.Among them,36 patients had hypertensive ICH and 3 had ICH due to ruptured intracranial aneurysms.(1)The time interval from head CT to super-resolution TCS imaging ranged from 0.33 to 23.73 h,with a median interval of 5.00(3.00,9.18)h.A total of 10 bleeding sites and 119 hematomas were detected by head CT,among which intraventricular hematomas accounted for 27.7％(33/119),subarachnoid hematomas for 19.3％(23/119),basal ganglia hematomas for 17.6％(21/119),frontal lobe hematomas for 9.2％(11/119),temporal lobe hematomas for 9.2％(11/119),thalamic hematomas for 5.9％(7/119),parieto-occipital lobe hematomas for 4.2％(5/119),cerebellar hematomas for 3.4％(4/119),insular lobe hematomas for 2.5％(3/119),and centrum semiovale hematomas for 0.8％(1/119).The detection rate of ICH by super-resolution TCS imaging was 89.92％,with 107 true positives,18 false positives,0 true negatives,and 12 false negatives.The sensitivity for hematoma detection was 89.92％,the accuracy was 78.10％,and the positive predictive value was 85.60％.Super-resolution TCS imaging showed high detection accuracy for intraventricular,subarachnoid space and basal ganglia hemorrhages,at 94.12％,95.83％and 95.45％,respectively,whereas the detection accuracy was relatively low for hematomas in the parieto-occipital lobe,insular lobe,and centrum semiovale,at 42.86％,40.00％,and 0,respectively.(2)After excluding hematomas that could not be reliably quantified by super-resolution TCS imaging,41 hematomas with well-defined boundaries suitable for measurement were retained.Inter-observer consistency evaluation demonstrated good consistency(ICC=0.880,95％CI 0.787-0.934)for measurements of hematoma mediolateral diameter and excellent consistency(ICC=0.903,95％CI 0.825-0.947)for measurements of anteroposterior diameter using super-resolution TCS imaging,all date points fell within the 95％LOA.The maximum anteroposterior diameters of 41 hematomas measured by super-resolution TCS imaging and cranial CT were(4.36±1.21)cm and(3.99±1.47)cm,respectively,and the maximum mediolateral diameters were(3.80±1.19)cm and(3.40±1.24)cm,respectively;the differences were statistically significant(both P＜0.01).The median relative errors between the maximum anteroposterior diameters and maximum mediolateral diameters measured by super-resolution TCS imaging and cranial CT were 11.6％(6.3％,28.1％)and 15.4％(7.8％,30.4％),respectively.The consistency evaluation showed that the intraclass ICC between the maximum anteroposterior diameters of hematomas measured by super-resolution TCS imaging and cranial CT was 0.843(95％CI 0.659-0.922),and the ICC for the maximum mediolateral diameter was 0.827(95％CI 0.564-0.921),both indicating good consistency.Bland-Altman analysis showed that compared with cranial CT,measurements by super-resolution TCS imaging had a systematic overestimation of the maximum anteroposterior diameter and maximum mediolateral diameter of hematomas(the mean differences were[0.37±0.68]cm and[0.41±0.62]cm,respectively),while most data points fell within the 95％LOA.Pearson correlation analysis showed that the maximum anteroposterior diameters measured by super-resolution TCS imaging and cranial CT had a strong positive correlation(|r|=0.867,P＜0.01),and the maximum mediolateral diameters also had a strong positive correlation(|r|=0.891,P＜0.01).Conclusions Super-resolution TCS imaging demonstrates high sensitivity and accuracy in the detection of ICH and shows good reliability in the measurement of hematoma diameters.It can serve as a low-cost,efficient bedside supplementary monitoring tool for hematoma surveillance in critically ill patients to assist clinical decision-making and condition assessment.However,it can not currently replace head CT as the standard for diagnosis.