Class: VAExtractor

Inherits:
Object
  • Object
show all
Defined in:
lib/vaextractor.rb

Constant Summary collapse

@@varegex = 
/(\s|^|~|:)(20|3E|E)\/\s*(\d+)\s*([+|-])*\s*(\d)*|(HM|CF|LP|NLP)(\W+(@|at|x)*\s*((\d+)(\s*'|\s*"|\s*in|\s*ft|\s*feet)*|face)*|$)/
@@snellenlevels = 
[10,15,20,25,30,40,50,60,70,80,100,125,150,200,250,300,400,600,800]
@@validtokens = 
{"OD" => "OD", "RE" => "OD", "RIGHT" => "OD", "R" => "OD", 
"L" => "OS", "OS" => "OS", "LE" => "OS", "LEFT" => "OS", 
"BOTH" => "OU", "BE" => "OU", "OU" => "OU", "BILATERAL" => "OU"}

Instance Method Summary collapse

Constructor Details

#initialize ⇒ VAExtractor

Returns a new instance of VAExtractor.



27
28
29
30
31
32
 
# File 'lib/vaextractor.rb', line 27

def initialize
	@usedsnellen = {}
	@@snellenlevels.each do |k|
		@usedsnellen[k] = -Math.log(20.0 / k) / Math.log(10.0)
	end
end

Instance Method Details

#aligntokens(s) ⇒ Object 



34
35
36
37
38
39
40
41
42
43
44
45
 
# File 'lib/vaextractor.rb', line 34

def aligntokens(s)
	arr = Textoken(s).tokens
	ret = []
	lasti = 0
	arr.each do |w|
		subs = s[lasti, s.size]
		i = subs.index(w)
		ret.push [w, lasti + i]
		lasti += i
	end
	return ret
end

#extract(rawtext) ⇒ Object 



143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
 
# File 'lib/vaextractor.rb', line 143

def extract(rawtext)
	lines = rawtext.split("\n")
	debug = false

	rfound = false
	lfound = false
	found = false
	vas = {"OD" => [], "OS" => []}
	alreadychecked = {}
	debugtxt = ""
	for i in (0...lines.count)
		next if lines[i].strip=~ /^IOP/ or lines[i].strip =~ /^Ta\s/ or lines[i].strip =~ /^Tp/i
		next if alreadychecked.has_key?(i)
		arr = lines[i].scan(@@varegex)
		lines[i].enum_for(:scan, @@varegex).each do |val|
			#puts "============================================="
			debugtxt += "NEW VA DETECTED\n"
			debugtxt += "#{lines[i-1]}\n"
			debugtxt += "#{lines[i]}\n"
			debugtxt += "#{lines[i+1]}\n"
			debugtxt += "#{val}\n"
			val.shift
			next if val[3] != nil  and val[3].to_i >= 5
			pos = Regexp.last_match.begin(0)
			tokens = aligntokens(lines[i])
			lat,debughash = findlaterality(pos, tokens, lines[i])
			debugtxt += "#{pos}\n"
			debugtxt += "#{tokens}\n"
			debugtxt += "#{debughash}\n"
			#p lat
			if lat == "OU"
				vas["OD"].push [val, i, 5]
				vas["OS"].push [val, i, 5]
			elsif lat != nil
				vas[lat].push  [val, i, 5]
			elsif lat == nil and vas["OD"].count == 0 and vas["OS"].count == 0
				lat = searchpriorlines(lines[i-3..i-1].reverse)
				if lat == nil and vas["OD"].count == 0 and vas["OS"].count == 0
					# most likely the VAs are either two in one line OD/OS or on two consecutive lines
					found = false
					arr2 = lines[i+1].scan(@@varegex)
					if arr2.count > 0 and arr.count > 0
						arr.each do |row|
							row.shift
							next if row[0] != nil  and row[3].to_i >= 5
							vas["OD"].push [row, i, 0]
						end
						arr2.each do |row|
							row.shift
							next if row[0] != nil  and row[3].to_i >= 5
							vas["OS"].push [row, i+1, 0]
						end
						alreadychecked[i+1] = 1
						found = true
					elsif arr.count == 2
						arr[0].shift
						arr[1].shift
						next if arr[0] != nil  and arr[0][3].to_i >= 5
						next if arr[1] != nil  and arr[1][3].to_i >= 5
						vas["OD"].push [arr[0], i, 0]
						vas["OS"].push [arr[1], i, 0]
						found = true
					end
					if not found 
						# worst case scenario, count up all the occurences of r/l and then take highest occuring freq
						lat = runentirefreq(rawtext)
						if lat == nil
							#puts "ERROR: Laterality not found for #{val}"
							raise ErrorLateralityNotFound
						else
							vas[lat].push [val, i, 0]
						end
					end
				else
					vas[lat].push  [val, i, 3]
				end
				#exit
			end
		end
	end

	if vas["OD"].count == 0 and vas["OS"].count == 0
		#puts "ERROR: No valid visual acuities found"
		return  {:RE => nil, :LE => nil, :RElogmar => nil, :LElogmar => nil}
	else
		bcva = {"OD" => nil, "OS" => nil}
		puts "=================NEW PT" if debug
		puts rawtext if debug
		puts "===DEBUG" if debug
		puts debugtxt if debug
		puts "===OD" if debug

		vas["OD"].each do |varr,line,priority|
			puts "new va" if debug
			p varr if debug
			p lines[line] if debug
			p priority if debug
			lva = logmar(varr)
			p lva if debug
			if bcva["OD"] == nil or (bcva["OD"][0] <= priority  and lva[0] < bcva["OD"][1])
				bcva["OD"] = [priority, lva[0].round(4), lva[1]]
			end
		end
		puts "===OS" if debug
		vas["OS"].each do |varr,line,priority|
			puts "new va" if debug
			p varr if debug
			p lines[line] if debug
			p priority if debug
			lva = logmar(varr)
			p lva if debug
			if bcva["OS"] == nil or (bcva["OS"][0] <= priority  and lva[0] < bcva["OS"][1])
				bcva["OS"] = [priority, lva[0].round(4), lva[1]]
			end
		end
		bcva["OD"] = [nil, nil, [nil,nil,nil,nil]] if bcva["OD"] == nil
		bcva["OS"] = [nil, nil, [nil,nil,nil,nil]] if bcva["OS"] == nil
		return  {:RE => bcva["OD"][2], :LE => bcva["OS"][2], :RElogmar => bcva["OD"][1], :LElogmar => bcva["OS"][1]}
	end
end

#findlaterality(pos, tokens, linestr) ⇒ Object 



109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
 
# File 'lib/vaextractor.rb', line 109

def findlaterality(pos, tokens, linestr)
	walls = {"." => 10, "!" => 10, "?" => 10, "," => 5, "and" => 5}
	answers = {}
	debug = ""
	revtoken = {}
	tokens.each do |w, i|
		revtoken[i] = w
	end
	tokens.each do |w, i|
		w = w.upcase
		if @@validtokens.has_key?(w)
			score = 0
			l = i
			r = pos
			l = pos if i > pos
			r = i if i > pos
			(l...r).each do |j|
				next if not revtoken.has_key?(j)
				score += walls[revtoken[j]] if walls.has_key?(revtoken[j])
			end
			if not answers.has_key?(score)
				answers[score] = []
			end
			answers[score].push [@@validtokens[w], (i-pos).abs]
		end
	end
	return nil if answers.keys.count == 0

	bestscore = answers.sort_by {|k, v| k}.first.last
	sorted = bestscore.sort_by {|r| r[1]}
	return sorted.first[0], answers
end

#logmar(va) ⇒ Object 



62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
 
# File 'lib/vaextractor.rb', line 62

def logmar(va)
	if va[0] == "20"
		manual = @usedsnellen[va[1].to_i]
		if va[2] == "+"
			va[3] = "1" if va[3] == nil
			denom = @@snellenlevels.index(va[1].to_i)
			denom -= 1
			denom = @usedsnellen[@@snellenlevels[denom]]
			manual = 1.0 * va[3].to_i * (denom - manual) / 5.0 + manual
		elsif va[2] == "-"
			va[3] = "1" if va[3] == nil
			denom = @@snellenlevels.index(va[1].to_i)
			denom += 1
			denom = @usedsnellen[@@snellenlevels[denom]]
			manual = 1.0 * va[3].to_i * (denom - manual) / 5.0 + manual
		end
		return manual, va[0...4]
	elsif va[0] == "3E" or va[0] == "3" or va[0] == "E"
		denom = va[1].to_i
		return -Math.log(3.0 / denom) / Math.log(10.0), va[0...4]
	elsif va[4] == "CF"
		return 2.0, [va[4],va[8],va[9],nil]
	elsif va[4] == "HM"
		return 2.4, [va[4],va[8],va[9],nil]
	elsif va[4] == "LP"
		return 2.7, [va[4],va[8],va[9],nil]
	elsif va[4] == "NLP"
		return 3.0, [va[4],va[8],va[9],nil]
	end
	return nil, nil
end

#runentirefreq(rawtext) ⇒ Object 



47
48
49
50
51
52
53
54
55
56
57
58
59
 
# File 'lib/vaextractor.rb', line 47

def runentirefreq(rawtext)
	tokens = aligntokens(rawtext)
	scores = Hash.new(0)
	tokens.each do |w, i|
		w = w.upcase
		if @@validtokens.has_key?(w) and @@validtokens[w] != "OU"
			scores[@@validtokens[w]] += 1
		end
	end
	return nil if scores.keys.count == 0
	#p scores
	return scores.sort_by {|k,v| v}.last.first
end

#searchpriorlines(lines) ⇒ Object 



94
95
96
97
98
99
100
101
102
103
104
105
106
 
# File 'lib/vaextractor.rb', line 94

def searchpriorlines(lines)
	lines.each do |l|
		return nil if l.strip == ""
		arr = Textoken(l).tokens
		arr.each do |w|
			if @@validtokens.has_key?(w.upcase)
				next if @@validtokens[w.upcase] == "OU"
				return @@validtokens[w.upcase]
			end
		end
	end
	return nil
end