dbdicom 0.3.0__py3-none-any.whl → 0.3.2__py3-none-any.whl

dbdicom/api.py

@@ -1,5 +1,4 @@
 
-import numpy as np
 import vreg
 
 from dbdicom.dbd import DataBaseDicom
@@ -16,6 +15,15 @@ def open(path:str) -> DataBaseDicom:
     """
     return DataBaseDicom(path)
 
+def to_json(path):
+    """Summarise the contents of the DICOM folder in a json file
+
+    Args:
+        path (str): path to the DICOM folder
+    """
+    dbd = open(path)
+    dbd.close()   
+
 def print(path):
     """Print the contents of the DICOM folder
 
@@ -24,6 +32,7 @@ def print(path):
     """
     dbd = open(path)
     dbd.print()
+    dbd.close()
 
 
 def summary(path) -> dict:
@@ -36,7 +45,24 @@ def summary(path) -> dict:
         dict: Nested dictionary with summary information on the database.
     """
     dbd = open(path)
-    return dbd.summary()
+    s = dbd.summary()
+    dbd.close()
+    return s
+
+
+def tree(path) -> dict:
+    """Return the structure of the database as a dictionary tree.
+
+    Args:
+        path (str): path to the DICOM folder
+
+    Returns:
+        dict: Nested dictionary with summary information on the database.
+    """
+    dbd = open(path)
+    s = dbd.register
+    dbd.close()
+    return s
 
 
 def patients(path, name:str=None, contains:str=None, isin:list=None)->list:
@@ -56,7 +82,9 @@ def patients(path, name:str=None, contains:str=None, isin:list=None)->list:
         list: list of patients fulfilling the criteria.
     """
     dbd = open(path)
-    return dbd.patients(name, contains, isin)
+    p = dbd.patients(name, contains, isin)
+    dbd.close()
+    return p
 
 
 def studies(entity:str | list, name:str=None, contains:str=None, isin:list=None)->list:
@@ -79,10 +107,14 @@ def studies(entity:str | list, name:str=None, contains:str=None, isin:list=None)
     """
     if isinstance(entity, str): # path = folder
         dbd = open(entity)
-        return dbd.studies(entity, name, contains, isin)
+        s = dbd.studies(entity, name, contains, isin)
+        dbd.close()
+        return s
     elif len(entity)==2: # path = patient
         dbd = open(entity[0])
-        return dbd.studies(entity, name, contains, isin)
+        s = dbd.studies(entity, name, contains, isin)
+        dbd.close()
+        return s
     else:
         raise ValueError(
             "The path must be a folder or a 2-element list "
@@ -110,10 +142,14 @@ def series(entity:str | list, name:str=None, contains:str=None, isin:list=None)-
     """
     if isinstance(entity, str): # path = folder
         dbd = open(entity)
-        return dbd.series(entity, name, contains, isin)
+        s = dbd.series(entity, name, contains, isin)
+        dbd.close()
+        return s
     elif len(entity) in [2,3]:
         dbd = open(entity[0])
-        return dbd.series(entity, name, contains, isin)
+        s = dbd.series(entity, name, contains, isin)
+        dbd.close()
+        return s
     else:
         raise ValueError(
             "To retrieve a series, the entity must be a database, patient or study."
@@ -168,7 +204,9 @@ def volume(series:list, dims:list=None, multislice=False) -> vreg.Volume3D:
         vreg.Volume3D: vole read from the series.
     """
     dbd = open(series[0])
-    return dbd.volume(series, dims, multislice)
+    vol = dbd.volume(series, dims, multislice)
+    dbd.close()
+    return vol
 
 def write_volume(vol:vreg.Volume3D, series:list, ref:list=None, 
                  multislice=False):
@@ -200,6 +238,7 @@ def to_nifti(series:list, file:str, dims:list=None, multislice=False):
     """
     dbd = open(series[0])
     dbd.to_nifti(series, file, dims, multislice)
+    dbd.close()
 
 def from_nifti(file:str, series:list, ref:list=None, multislice=False):
     """Create a DICOM series from a nifti file.
@@ -232,7 +271,9 @@ def pixel_data(series:list, dims:list=None, include:list=None) -> tuple:
             return value.
     """
     dbd = open(series[0])
-    return dbd.pixel_data(series, dims, include)
+    array = dbd.pixel_data(series, dims, include)
+    dbd.close()
+    return array
 
 # write_pixel_data()
 # values()
@@ -255,7 +296,9 @@ def unique(pars:list, entity:list) -> dict:
         dict: dictionary with unique values for each attribute.
     """
     dbd = open(entity[0])
-    return dbd.unique(pars, entity)
+    u = dbd.unique(pars, entity)
+    dbd.close()
+    return u
 
 
 

dbdicom/database.py

@@ -0,0 +1,122 @@
+import os
+from tqdm import tqdm
+
+import numpy as np
+import pandas as pd
+import pydicom
+
+import dbdicom.utils.dcm4che as dcm4che
+import dbdicom.utils.files as filetools
+import dbdicom.dataset as dbdataset
+
+
+COLUMNS = [   
+    # Identifiers (unique)
+    'PatientID', 
+    'StudyInstanceUID', 
+    'SeriesInstanceUID', 
+    'SOPInstanceUID', 
+    # Human-readable identifiers (not unique)
+    'PatientName', 
+    'StudyDescription', 
+    'StudyDate', 
+    'SeriesDescription', 
+    'SeriesNumber', 
+    'InstanceNumber', 
+]
+
+def read(path):
+    files = filetools.all_files(path)
+    tags = COLUMNS + ['NumberOfFrames'] # + ['SOPClassUID']
+    array = []
+    dicom_files = []
+    for i, file in tqdm(enumerate(files), total=len(files), desc='Reading DICOM folder'):
+        try:
+            ds = pydicom.dcmread(file, force=True, specific_tags=tags+['Rows'])
+        except:
+            pass
+        else:
+            if isinstance(ds, pydicom.dataset.FileDataset):
+                if 'TransferSyntaxUID' in ds.file_meta:
+                    if not 'Rows' in ds: # Image only
+                        continue
+                    row = dbdataset.get_values(ds, tags)
+                    array.append(row)
+                    index = os.path.relpath(file, path)
+                    dicom_files.append(index) 
+    df = pd.DataFrame(array, index = dicom_files, columns = tags)
+    df = _multiframe_to_singleframe(path, df)
+    dbtree = _tree(df)
+    return dbtree
+
+
+def _multiframe_to_singleframe(path, df):
+    """Converts all multiframe files in the folder into single-frame files.
+    
+    Reads all the multi-frame files in the folder,
+    converts them to singleframe files, and delete the original multiframe file.
+    """
+    singleframe = df.NumberOfFrames.isnull() 
+    multiframe = singleframe == False
+    nr_multiframe = multiframe.sum()
+    if nr_multiframe != 0: 
+        for relpath in tqdm(df[multiframe].index.values, desc="Converting multiframe file " + relpath):
+            filepath = os.path.join(path, relpath)
+            singleframe_files = dcm4che.split_multiframe(filepath) 
+            if singleframe_files != []:            
+                # add the single frame files to the dataframe
+                dfnew = read(singleframe_files, df.columns, path)
+                df = pd.concat([df, dfnew])
+                # delete the original multiframe 
+                os.remove(filepath)
+            # drop the file also if the conversion has failed
+            df.drop(index=relpath, inplace=True)
+    df.drop('NumberOfFrames', axis=1, inplace=True)
+    return df
+
+
+def _tree(df):
+    # A human-readable summary tree
+
+    df.sort_values(['PatientID','StudyInstanceUID','SeriesNumber'], inplace=True)
+    df = df.fillna('None')
+    summary = []
+
+    for uid_patient in df.PatientID.unique():
+        df_patient = df[df.PatientID == uid_patient]
+        patient_name = df_patient.PatientName.values[0]
+        patient = {
+            'PatientName': patient_name,
+            'PatientID': uid_patient,
+            'studies': [],
+        }
+        summary.append(patient)
+        for uid_study in df_patient.StudyInstanceUID.unique():
+            df_study = df_patient[df_patient.StudyInstanceUID == uid_study]
+            study_desc = df_study.StudyDescription.values[0]
+            study_date = df_study.StudyDate.values[0]
+            study = {
+                'StudyDescription': study_desc,
+                'StudyDate': study_date,
+                'StudyInstanceUID': uid_study,
+                'series': [],
+            }
+            patient['studies'].append(study)
+            for uid_sery in df_study.SeriesInstanceUID.unique():
+                df_series = df_study[df_study.SeriesInstanceUID == uid_sery]
+                series_desc = df_series.SeriesDescription.values[0]
+                series_nr = int(df_series.SeriesNumber.values[0])
+                series = {
+                    'SeriesNumber': series_nr,
+                    'SeriesDescription': series_desc,
+                    'SeriesInstanceUID': uid_sery,
+                    'instances': {},
+                }
+                study['series'].append(series)
+                for uid_instance in df_series.SOPInstanceUID.unique():
+                    df_instance = df_series[df_series.SOPInstanceUID == uid_instance]
+                    instance_nr = int(df_instance.InstanceNumber.values[0])
+                    relpath = df_instance.index[0]
+                    series['instances'][instance_nr]=relpath
+
+    return summary

dbdicom/dataset.py

@@ -1,16 +1,18 @@
+# Test data
+# https://www.aliza-dicom-viewer.com/download/datasets
+
 import os
 from datetime import datetime
 import struct
 from tqdm import tqdm
 
 import numpy as np
-import pandas as pd
 import pydicom
 from pydicom.util.codify import code_file
 import pydicom.config
-from pydicom.dataset import Dataset
 import vreg
 
+
 import dbdicom.utils.image as image
 import dbdicom.utils.variables as variables
 from dbdicom.sop_classes import (
@@ -71,6 +73,8 @@ def new_dataset(sop_class):
         return xray_angiographic_image.default()
     if sop_class == 'UltrasoundMultiFrameImage':
         return ultrasound_multiframe_image.default()
+    if sop_class == 'ParametricMap':
+        return parametric_map.default()
     else:
         raise ValueError(
             f"DICOM class {sop_class} is not currently supported"
@@ -235,7 +239,7 @@ def codify(source_file, save_file, **kwargs):
     file.close()
 
 
-def read_data(files, tags, path=None, images_only=False):
+def read_data(files, tags, path=None, images_only=False): # obsolete??
 
     if np.isscalar(files):
         files = [files]
@@ -263,34 +267,6 @@ def read_data(files, tags, path=None, images_only=False):
 
 
 
-def read_dataframe(files, tags, path=None, images_only=False):
-    if np.isscalar(files):
-        files = [files]
-    if np.isscalar(tags):
-        tags = [tags]
-    array = []
-    dicom_files = []
-    for i, file in tqdm(enumerate(files), desc='Reading DICOM folder'):
-        try:
-            ds = pydicom.dcmread(file, force=True, specific_tags=tags+['Rows'])
-        except:
-            pass
-        else:
-            if isinstance(ds, pydicom.dataset.FileDataset):
-                if 'TransferSyntaxUID' in ds.file_meta:
-                    if images_only:
-                        if not 'Rows' in ds:
-                            continue
-                    row = get_values(ds, tags)
-                    array.append(row)
-                    if path is None:
-                        index = file
-                    else:
-                        index = os.path.relpath(file, path)
-                    dicom_files.append(index) 
-    df = pd.DataFrame(array, index = dicom_files, columns = tags)
-    return df
-
 
 def _add_new(ds, tag, value, VR='OW'):
     if not isinstance(tag, pydicom.tag.BaseTag):
@@ -583,7 +559,7 @@ def pixel_data(ds):
     try:
         array = ds.pixel_array
     except:
-        return None
+        raise ValueError("Dataset has no pixel data.")
     array = array.astype(np.float32)
     slope = float(getattr(ds, 'RescaleSlope', 1)) 
     intercept = float(getattr(ds, 'RescaleIntercept', 0)) 
@@ -592,7 +568,7 @@ def pixel_data(ds):
     return np.transpose(array)
 
 
-def set_pixel_data(ds, array, value_range=None):
+def set_pixel_data(ds, array):
     if array is None:
         raise ValueError('The pixel array cannot be set to an empty value.')
     
@@ -608,7 +584,7 @@ def set_pixel_data(ds, array, value_range=None):
     # if array.ndim >= 3: # remove spurious dimensions of 1
     #     array = np.squeeze(array) 
 
-    array = image.clip(array.astype(np.float32), value_range=value_range)
+    array = image.clip(array.astype(np.float32))
     array, slope, intercept = image.scale_to_range(array, ds.BitsAllocated)
     array = np.transpose(array)
 
@@ -632,6 +608,15 @@ def volume(ds, multislice=False):
 def set_volume(ds, volume:vreg.Volume3D, multislice=False):
     if volume is None:
         raise ValueError('The volume cannot be set to an empty value.')
+    try:
+        mod = SOPCLASSMODULE[ds.SOPClassUID]
+    except KeyError:
+        raise ValueError(
+            f"DICOM class {ds.SOPClassUID} is not currently supported."
+        )
+    if hasattr(mod, 'set_volume'):
+        return getattr(mod, 'set_volume')(ds, volume)
+    
     image = np.squeeze(volume.values)
     if image.ndim != 2:
         raise ValueError("Can only write 2D images to a dataset.")
@@ -712,41 +697,7 @@ def set_signal_type(ds, value):
 
 
 
+if __name__=='__main__':
 
-# def _initialize(ds, UID=None, ref=None): # ds is pydicom dataset
-
-#     # Date and Time of Creation
-#     dt = datetime.now()
-#     timeStr = dt.strftime('%H%M%S')  # long format with micro seconds
-
-#     ds.ContentDate = dt.strftime('%Y%m%d')
-#     ds.ContentTime = timeStr
-#     ds.AcquisitionDate = dt.strftime('%Y%m%d')
-#     ds.AcquisitionTime = timeStr
-#     ds.SeriesDate = dt.strftime('%Y%m%d')
-#     ds.SeriesTime = timeStr
-#     ds.InstanceCreationDate = dt.strftime('%Y%m%d')
-#     ds.InstanceCreationTime = timeStr
-
-#     if UID is not None:
-
-#         # overwrite UIDs
-#         ds.PatientID = UID[0]
-#         ds.StudyInstanceUID = UID[1]
-#         ds.SeriesInstanceUID = UID[2]
-#         ds.SOPInstanceUID = UID[3]
-
-#     if ref is not None: 
-
-#         # Series, Instance and Class for Reference
-#         refd_instance = Dataset()
-#         refd_instance.ReferencedSOPClassUID = ref.SOPClassUID
-#         refd_instance.ReferencedSOPInstanceUID = ref.SOPInstanceUID
-
-#         refd_series = Dataset()
-#         refd_series.ReferencedInstanceSequence = Sequence([refd_instance])
-#         refd_series.SeriesInstanceUID = ds.SeriesInstanceUID
-
-#         ds.ReferencedSeriesSequence = Sequence([refd_series])
-
-#     return ds
+    pass
+    #codify('C:\\Users\\md1spsx\\Documents\\f32bit.dcm', 'C:\\Users\\md1spsx\\Documents\\f32bit.py')