def train_model(self, UI=None):
if self.k <= 0 and UI is not None:
UI.msg = QMessageBox()
UI.msg.setIcon(QMessageBox.Warning)
UI.msg.setWindowTitle("Warning")
UI.msg.setText(
"Anda harus membagi data menggunakan k-fold terlebih dahulu")
UI.msg.setStandardButtons(QMessageBox.Ok)
UI.msg.show()
UI.statusBar().showMessage("Train and test failed")
return False
self.attrs, self.clfs, self.tfidfs = ([0 for i in range(self.k)
], [0 for i in range(self.k)],
[0 for i in range(self.k)])
el = QEventLoop()
for i in range(self.k):
if UI is not None:
UI.logOutput.append(f"Train tree {i + 1}")
params = {'i': i, 'remove_zero_tfidf': True, 'UI': UI}
worker = Worker(self.mltrain_fn, params)
worker.signals.result.connect(self.mltrain_result)
self.threadpool.start(worker)
self.threadpool.waitForDone()
el.processEvents()
el.exit()
if UI is not None:
UI.logOutput.append("Training completed")
return self.attrs
def _killProcesses(processes: List[Popen]):
if any(p.poll() is None for p in processes):
loop = QEventLoop()
QTimer.singleShot(videoSettings.VLC_SLEEP_TIME_LLmsJJ, loop.quit)
loop.processEvents()
for p in processes:
if p.poll():
continue
logger.debug(f'Killing process {p.pid}')
p.kill()
def load_hdf5_files(self, filenames):
n_files = len(filenames)
for f in filenames:
if not h5py.is_hdf5(f):
raise TypeError(f'{f} is not a valid HDF5 file.')
loop = QEventLoop()
progress = QProgressDialog("Loading location tables", None, 0, n_files,
self._parent)
plural = 'files' if n_files > 1 else 'file'
progress.setWindowTitle(f"Loading {n_files} location table {plural}")
progress.setModal(True)
progress.setMinimumDuration(0)
progress.setValue(0)
progress.show()
loop.processEvents()
for i, filename in enumerate(filenames):
status_message = f'loading {filename}'
progress.setLabelText(status_message)
progress.setValue(i)
loop.processEvents()
self.load_hdf5(filename)
progress.setValue(i + 1)
loop.processEvents()
progress.close()
def _load_json(self, filename):
"""
Loads all of the HDF5 files listed in the input JSON file.
Args:
filename: the fully-qualified name of a JSON file containing the names of HDF5 files produced by a
DecontaminatedSpectraCollection.
"""
if not os.path.isabs(filename):
raise ValueError(
"The filename must include the full (absolute) path.")
dir_name = os.path.dirname(filename)
with open(filename) as f:
decontaminated_spectra_collection_filenames = json.load(f)
for f in decontaminated_spectra_collection_filenames:
full_name = os.path.join(dir_name, 'data', f)
if not h5py.is_hdf5(full_name):
raise TypeError(f'{f} is not a valid HDF5 file.')
n_files = len(decontaminated_spectra_collection_filenames)
loop = QEventLoop()
progress = QProgressDialog(
"Loading decontaminated spectra collections", None, 0, n_files,
self._parent)
plural = 'files' if n_files > 1 else 'file'
progress.setWindowTitle(
f"Loading {n_files} decontaminated spectra {plural}")
progress.setModal(True)
progress.setMinimumDuration(0)
progress.setValue(0)
progress.show()
loop.processEvents()
for i, decontaminated_spectra_filename in enumerate(
decontaminated_spectra_collection_filenames):
status_message = f'loading {decontaminated_spectra_filename}'
progress.setLabelText(status_message)
progress.setValue(i)
loop.processEvents()
full_name = os.path.join(dir_name, 'data',
decontaminated_spectra_filename)
self._load_hdf5(full_name)
progress.setValue(i + 1)
loop.processEvents()
progress.close()
class MainWindow(QMainWindow):
donesignal = pyqtSignal(tuple)
def __init__(self):
super(MainWindow, self).__init__()
self.ui = Ui_MainWindow()
self.ui.setupUi(self)
self.info = {}
self.collected_data = []
self.progress = 0 #actual img
self.target = 20 #img to evaluate
if debug:
self.target = 20
#self.ui.img = DicomWidget(self) #lavorare qua per piazzarlo
self.ui.pushButtonStart.clicked.connect(lambda: self.load_at_start())
if debug:
## Set the imgs
pix_not = QPixmap(r'ui\img_not_sufficient.png').scaled(50, 50)
self.ui.labelInsufficientImg.setPixmap(pix_not)
pix_low = QPixmap(r'ui\img_low.png').scaled(50, 50)
self.ui.labelLowImg.setPixmap(pix_low)
pix_med = QPixmap(r'ui\img_medium.png').scaled(50, 50)
self.ui.labelMediumImg.setPixmap(pix_med)
pix_ok = QPixmap(r'ui\img_max.png').scaled(50, 50)
self.ui.labelMaxImg.setPixmap(pix_ok)
## Set the imgs
pix_not = QPixmap(
r'C:\Users\IEO5572\PycharmProjects\qt_test_image\ui\img_not_sufficient.png'
).scaled(50, 50)
self.ui.labelInsufficientImg.setPixmap(pix_not)
pix_low = QPixmap(
r'C:\Users\IEO5572\PycharmProjects\qt_test_image\ui\img_low.png'
).scaled(50, 50)
self.ui.labelLowImg.setPixmap(pix_low)
pix_med = QPixmap(
r'C:\Users\IEO5572\PycharmProjects\qt_test_image\ui\img_medium.png'
).scaled(50, 50)
self.ui.labelMediumImg.setPixmap(pix_med)
pix_ok = QPixmap(
r'C:\Users\IEO5572\PycharmProjects\qt_test_image\ui\img_max.png'
).scaled(50, 50)
self.ui.labelMaxImg.setPixmap(pix_ok)
self.ui.pushButtonInsufficient.clicked.connect(lambda: self.vote(0))
self.ui.pushButtonLow.clicked.connect(lambda: self.vote(1))
self.ui.pushButtonMedium.clicked.connect(lambda: self.vote(2))
self.ui.pushButtonMax.clicked.connect(lambda: self.vote(3))
self.pool = QThreadPool.globalInstance()
self.pool.setMaxThreadCount(4) # Use up to 8 threads
#scroll_area = QtWidgets.QScrollArea()
#scroll_area.setWidget(self.pix_label)
def closeEvent(self, event):
"""Save collected data on exit"""
df = pd.DataFrame.from_dict(self.collected_data)
#df["CONTRASTO"]=self.contrast
phys_name = self.ui.plainTextPhysSign.toPlainText()
df.to_csv(
os.path.join(
r"\\portale.ieo.it\portale\Aree_Dipartimentali\G_RAD\2- G_RAD - ALTRE CARTELLE\MDC",
f"collected_data_{phys_name}.csv"))
print(f"[INFO] Output saved to collected_data.csv")
can_exit = True
if can_exit:
event.accept() # let the window close
else:
event.ignore()
def vote(self, rate):
"""This function add to info rate and physician name"""
phys_name = self.ui.plainTextPhysSign.toPlainText()
self.info["rate"] = rate
self.info["Physician"] = phys_name
self.info["contrast"] = self.contrast[self.progress]
self.collected_data.append(self.info)
self.progress += 1
self.ui.progressBar.setProperty("value", self.progress * 5)
self.magic()
def update_coordinates(self):
self.ui.img.update_image()
#mostrare le info di visualizzazione
try:
self.ui.infoLabel.setText(
f"Series: {self.header.SeriesDescription} \nWidth:\t {self.ui.img.window_width}\nCenter:\t {self.ui.img.window_center}"
)
except Exception as e:
print(e)
self.ui.infoLabel.setText(
f"Width:\t {self.ui.img.window_width}\nCenter:\t {self.ui.img.window_center}"
)
def cv2QtImg(self, path):
cv_img = cv2.imread(path)
rgb_image = cv2.cvtColor(cv_img, cv2.COLOR_BGR2RGB)
h, w, ch = rgb_image.shape
bytes_per_line = ch * w
convert_to_Qt_format = QtGui.QImage(rgb_image.data, w, h,
bytes_per_line,
QtGui.QImage.Format_RGB888)
#p = convert_to_Qt_format.scaled(self.display_width, self.display_height, Qt.KeepAspectRatio)
return QPixmap.fromImage(convert_to_Qt_format)
#LAVORARE QUA AGGIUNGERE MODIFICA CONTRASTO IMMAGINI E CARICARE DICOM
def load_at_start(self):
self.ui.pushButtonStart.setProperty("text", "Attendi")
#load all target images
self.studies = []
self.study_headers = []
self.study_mean_MA = []
self.contrast = []
self.patients = []
for i in range(self.target): #three images for security
print(f"[INFO] loading {i} of {self.target} images..")
self._file_name = self.sample()
self.patients.append(self._file_name)
#data,header,mean_mA=load_patient(self._file_name)
#self.studies.append(data)
#self.study_headers.append(header)
#self.study_mean_MA.append(mean_mA)
###SEPARATE THREAD APPROACH
self.wait_dialog()
self.runAll()
print("FINE", len(self.studies))
print(f"[INFO] loading completed in {time.time()-start_time} s")
self.ui.progressBar.setProperty("value", self.progress * 5)
self.ui.pushButtonStart.setVisible(False)
self.ui.pushButtonStart.setEnabled(False)
#self.magic()
def runAll(self):
self.eventloop = QEventLoop()
self.pool = QThreadPool.globalInstance()
self.pool.setMaxThreadCount(4)
self.donesignal.connect(self.add_patient)
for j in range(5):
for i in range(4):
worker = AWorker(target=self.load_patient,
args=(self.patients[4 * i + j], ))
self.pool.start(worker)
self.pool.waitForDone()
self.eventloop.processEvents()
self.eventloop.exit()
# Now your data should have been appended to your list
def load_patient(self, path):
print("[INFO] Loading patient data..")
if "NEW" in path:
contrast_type = "new"
elif "OLD" in path:
contrast_type = "old"
imgs = glob.glob(os.path.join(path, "*.dcm"))
#
try:
data, header, mean_mA = DicomData.from_files(imgs)
except:
print("Could not load files")
#emitting output
#self.data.emit(data)
#self.header.emit(header)
#self.meanMa.emit(mean_mA)
self.donesignal.emit((data, header, mean_mA, contrast_type))
self.add_patient((data, header, mean_mA, contrast_type))
return data, header, mean_mA
def add_patient(self, data_tuple):
print(f"[INFO] Adding patient..")
act = self.ui.progressBar.value()
self.ui.progressBar.setProperty("value", act + 5)
data, header, mean_mA, contrast_type = data_tuple
self.studies.append(data)
self.study_headers.append(header)
self.study_mean_MA.append(mean_mA)
self.contrast.append(contrast_type)
self.magic()
def wait_dialog(self):
msg = QMessageBox()
msg.setIcon(QMessageBox.Information)
msg.setText(
"Caricamento esami in corso, per favore aspetta qualche minuto, l'immagine compararirà automaticamente."
)
msg.setWindowTitle("Caricamento")
msg.setStandardButtons(QMessageBox.Ok | QMessageBox.Cancel)
retval = msg.exec_()
def showdialog(self):
msg = QMessageBox()
msg.setIcon(QMessageBox.Information)
msg.setText("Fine degli esami caricati")
msg.setInformativeText(
"Puoi chiudere questo box e il programma cliccando sulla X")
msg.setWindowTitle("Fine esami")
msg.setDetailedText(
"Ricorda di inserire il tuo nome nel box in alto a sinistra prima di chiudere."
)
msg.setStandardButtons(QMessageBox.Ok | QMessageBox.Cancel)
retval = msg.exec_()
def magic(self):
"""This function show the actual img"""
print(self.progress)
if self.progress < self.target:
self.header = self.study_headers[self.progress]
self.get_header_info(self.study_mean_MA[self.progress])
self.ui.img.data = self.studies[self.progress]
if self.progress == 0: # if it's the first img set the width and the center to optimal values
self.ui.img.window_center = 40.
self.ui.img.window_width = 400.
#self.setWindowTitle("IEO Dicom Viewer")
#self.file_list = imgs
self.ui.img.index = 0
self.ui.img.update_image()
else:
##DIALOG
self.showdialog()
def magic_old(self):
self._file_name = self.sample()
# #list files
imgs = glob.glob(os.path.join(self._file_name, "*.dcm"))
#
try:
self._file_name = self.sample()
# #list files
imgs = glob.glob(os.path.join(self._file_name, "*.dcm"))
data, header, mean_mA = DicomData.from_files(imgs)
self.header = header
self.get_header_info(mean_mA)
self.ui.img.data = data
if self.progress == 0: #if it's the first img set the width and the center to optimal values
self.ui.img.window_center = 40.
self.ui.img.window_width = 400.
self.setWindowTitle("IEO Dicom Viewer")
self.file_list = imgs
self.ui.img.index = 0
self.ui.img.update_image()
except BaseException as exc:
print(exc)
self.ui.img.data = None
self.setWindowTitle("IEO Dicom Viewer: No image")
def sample(self):
"""Pic a random number and choose newest or older images"""
r = randint(0, 1)
if r:
dir = NEW_DIR
self.info["contrast_protocol"] = "new"
else:
dir = OLD_DIR
self.info["contrast_protocol"] = "old"
subjects = os.listdir(dir)
rnd_sub_idx = randint(0, len(subjects) - 1)
return os.path.join(dir, subjects[rnd_sub_idx])
def get_header_info(self, mean_mA):
test = self.header
info = None
try:
info = {
"name": test.PatientName,
"sex": test.PatientSex,
"age": test.PatientAge,
"birthday": test.PatientBirthDate,
"patient_ID": test.PatientID,
"height": 0,
"weight": 0,
# study information
"study_date": test.StudyDate,
"study_time": test.StudyTime,
"study_id": test.StudyID,
"study_description": test.StudyDescription,
"series_description": test.SeriesDescription,
"model": test.ManufacturerModelName,
"protocol": test.ProtocolName,
# Other INFO
"kVp": float(test.KVP),
# "mA": float(test.XRayTubeCurrent), # QUA CAMBIA PER OGNI FETTA, FORSE ACQUISIZIONE?
"mA": mean_mA,
"contrast_agent": test.ContrastBolusAgent,
"filter_type": test.FilterType,
"convKernel": test.ConvolutionKernel,
}
try:
info["height"] = test.PatientSize
info["weight"] = test.PatientWeight
except:
print(f"[WARNING] Couldn't update PatientSize")
except Exception as e:
print(e)
print(
f"[INFO] An error occuered with {test.PatientName}. Saving your results"
)
self.close()
self.info = info
return info
